Professional Documents
Culture Documents
Properties of High Volume Fly Ash Concrete: Hsien-Hsin Hung
Properties of High Volume Fly Ash Concrete: Hsien-Hsin Hung
Hsien-Hsin Hung
May 1997
"PROPERTIES OF HIGH VOLUME FLY ASH CONCRETE"
SUMMARY
This thesis presents a detailed investigation on the engineering properties and microstructural
characteristics of concrete containing a high volume of fly ash ( HVFA ). The purpose of the
project is to evaluate the concept of using relatively large volumes of fly ash in nonnal
portland cement concrete, and hence enhance the beneficial use of fly ash in ,"alue-added
products and construction.
A total of eight concrete mixtures with and ,\ithout fly ash was investigated. The
proportion of fly ash in all the HVFA concrete mixtures varied from 50 to 80 % by \yeight of
the cementitious materials, with a constant water-to-cementitious ratio of 0.40 for all the
mixtures. A high degree of workability was maintained by the use of a superplasticizer. To
optimize the pozzolanic activity in the HVFA concrete, silica fume was used in some of the
mixes. The total cementitious materials content was kept constant at 350 kg/m 3 and 450
3
kg/m respectively. The influence of the different replacement materials and two curing
regimes was studied.
The study consisted of two parts. The first part is an ex.1ensive study of the
engineering properties such as strength development, modulus of elasticity, ultrasonic pulse
velocity, swelling, and drying shrinkage at various ages up to 18 months. The depth of
carbonation of HVFA concrete under different curing regimes was also investigated. A study
of the microstructure of HVF A concretes forms the second part of the investigation. Pore
structure, air permeability and water absorption of HVF A concretes "ith different
replacement mixtures were studied. A detailed discussion dealing with the change of the
morphological phase under different curing regimes is also presented.
The results show that HVF A concretes exhibit excellent mechanical properties ",ith
good long-term strength development. Compressive strength in the range of 40 to 60 MPa
"as achieved for all the HVFA concretes at the age of 90 days. The d)namic modulus of
elasticity reached values of the order of 55 GPa at 90 days. Under similar conditions,
concretes made \\ ith both fly ash and silica fume had engineering properties which were as
good as those made with cement replaced by fly ash alone. The use of fly ash to replace both
cement and sand has the advantage of mobilizing and combining the benefits and effects of
both separate replacements. The HVFA concretes also have low permeability and exhibit
good potential characteristics to resist water penetration.
Reduction in the volume of large pores was observed with the progress of the
pozzolanic reaction. Higher HVFA concrete strength was generally associated \\ith a lower
volume of large pores in the concrete. A decrease in the levels of calcium hydroxide was seen
with progressive water curing and age in all the HVFA concretes, providing evidence of
continued pozzolanic reactivity of the fly ashes.
Various empirical relationships and design equations are presented and conclusions
are drawn at the end of each part. It is recommended that further research is required to
determine the influence on HVF A concretes of extreme curing conditions such as high or low
temperature and low moisture availability, and to improve the early strength properties of the
HVFA concretes.
II
ACKNOWLEDGEMENTS
The author would like to express heartfelt gratitude to Professor R. N. Swamy, who
gave kindly guidance and advice at all stages of the study, offered many helpful
suggestions, and comments during this investigation. Many thanks are offered to
Him and Mrs. Swamy for their thoughtful and frequent concern about the life of the
author's family during the course of this study.
The author also wishes to acknowledge the assistance of the technical staff, in
particular Mr. N. Morton, Mr. A. Smart, of the Department of Mechanical
Engineering; and also from the Department of Civil and Structural Engineering, and
the Department of Engineering Materials.
The author is indebted to his wife, Chao-Chih, for her most encouragement,
and patient understanding throughout the research period, and the continuous
concern and constant support from his sisters and brothers.
III
Table of Contents
SUMMARy ............................................................................................................. i
ACKNOWLEDGEMENTS ..................................................................................... iii
T ABLES OF CONTENTS ..................................................................................... .iv
LIST OF TABLES .................................................................................................. ix
LIST OF FIGURES ............................................................................................... xiii
NOTATIONS AND ABBREVIATIONS ............................................................ xviii
IV
2-7-2. Sulphate resistance of concrete ................................................................. 26
2-7-3. Expansion due to alkali-aggregate reaction .............................................. 27
2-7-4. Carbonation .............................................................................................. 28
2-7-5. Chloride penetration ............................................ ..................................... 28
2-8. INFLUENCE OF SILICA FUME IN CONCRETE ......................................... 30
2-9. HIGH STRENGTH / HIGH PERFORMANCE CONCRETE AND HIGH
VOLUME FLY ASH CONCRETE ....................... '" ....................................... 31
\1
5-7-2. Calcium hydroxide (CH) ...................................................... ................. 1-16
5-7-3. Calcium sulfoalumimate hydrates ........................................................... 1-17
5-7-4. Carbonates ............................................................................................. 1-18
HI
curing ....................................................................................................... 208
7-5-2. Influence of air curing on X-ray diffraction ............................................ 217
7~6. CONCLUSION ............................................................................................. 220
viii
List of Tables
Table 3-7 High volume fly ash concrete workability ............................................... .45
Table 3-8 High volume fly ash concrete test details ................................................ .48
Table 4-1 Compressive strength ofHVF A concrete under water curing
(Unit: MPa ) ......................................................................................... 55
Table 4-2 Rate of gain of compressive strength as a % of 28 days strength
(Water curing) (Unit: % ) ................................................................... 56
Table 4-3 Compressive strength ofHVFA concrete under air curing
(Unit: MPa ) ......................................................................................... 58
Table 4-4 Rate of gain of compressive strength as a % of28 days strength
( Air curing) ( Unit: % ) ........................................................................ 59
3
Table 4-5 Influence of air curing on compressive strength-350 kglm binder
content ( Unit : % ) ................................................................................ 60
3
Table 4-6 Influence of air curing on compressive strength-450 kglm binder
content (Unit: %) ................................................................................ 60
Table 4-7 Flexural strength ofHVF A concrete under water curing
( Unit: MPa ) ......................................................................................... 63
Table 4-8 Rate of gain of flexural strength as a 010 of28 day strength
(Water curing) ( Unit: % ) ................................................................... 64
Table 4-9 Flexural strength of HVF A concrete under air curing ( Unit : MPa ) ........ 65
Table 4-10 Rate of gain of flexural strength as a % of 28 days strength
IX
( Air curing) (Unit : %) ......................................................................... 66
Table 4-11 Influence of air curing on flexural strength-350 kg/m3 binder content
(Unit: % ) ............................................................................................. 67
Table 4-12 Influence of air curing on flexural strength-450 kg/m3 binder content
(Unit: % ) ............................................................................................. 67
Table 4-13 Strength development of normal OPC concrete ..................................... 72
Table 4-14 Swelling strain ofHVFA concrete (Water curing)
6
(Unit: 10- mlm) .................................................................................. 79
Table 4-15 Drying shrinkage strain of HVF A concrete ( air curing)
(Unit: 10-6 mlm) ................................................................................... 84
Table 4-16 Typical % of meso pore of high volume fly ash concrete under air
curing ( Unit: % ) .................................................................................. 88
Table 4-17 Comparison between measured and predicted drying shrinkage
6
(Unit: 10- mlm) ................................................................................... 93
Table 4-18 Comparison of drying shrinkage with published data ............................. 94
Table 4-19 Dynamic modulus of elasticity of HVF A concrete ( Water curing)
(Unit: GPa) ......................................................................................... 96
Table 4-20 Dynamic modulus of elasticity ofHVFA concrete ( Air curing)
( Unit : GPa ) ......................................................................................... 96
Table 4-21 Influence of air curing on dynamic modulus of elasticity ( Unit: % ) ... 100
Table 4-22 Dynamic modulus of elasticity of normal OPC concrete ....................... 101
Table 4-23 Rate of gain of modulus as a % of28 day modulus (Water curing)
(Unit:%) ........................................................................................... 102
Table 4-24 Rate of gain of modulus as a % of 28 day modulus ( Air curing)
( Unit: % ) ........................................................................................... 102
Table 4-25 Comparison of predicting modulus of elasticity and published data under
water curing ( 28 days) ........................................................................ 107
Table 4-26 Comparison of predicting modulus elasticity and published data under
water curing ( 504 days) ...................................................................... 107
Table 4-27 Ultrasonic pulse velocity ofHVFA concrete (Water curing)
( Unit: kmlsec ) ................................................................................... 110
Table 4-28 Ultrasonic pulse velocity ofHVFA concrete (Air curing)
( Unit: kmlsec ) .................................................................................. III
Table 4-29 Influence of air curing on pulse velocity ( Unit: % ) ........................... 112
x
Table 4-30 Ultrasonic pulse velocity of normal OPC concrete ............................... 115
Table 4-31 Rate of gain of velocity as a % of28 day pulse velocity (Water curing)
( Unit: % ) ........................................................................................... 116
Table 4-32 Rate of gain of velocity as a % of28 day pulse velocity ( Air curing)
( Unit: % ) ........................................................................................... 116
Table 4-33 Carbonation depth - 350 kg/m3 mixes ( Unit: mm ) ............................. 120
3
Table 4-34 Carbonation depth - 450 kg/m mixes (Unit: mm ) ............................. 122
Table 4-35 Estimated parameters of equation C = a + KJi (t = sec ) .................. 126
Table 4-36 Comparison of carbonation depths with published data ........................ 130
Table 5-1 Classification of pore sizes in hydrated cement pastes ............................ 137
Table 6-1 A typical output ofMicromeritics pore size analyses ............................. 154
Table 7-1 Total accessible pore volume ofOPC and HVFA concrete
(Water curing) (Unit: mL/g) ............................................................ 168
Table 7-2 Total accessible pore volume of OPC and HVF A concrete
(Air curing) (Unit: mL/g ) ................................................................. 168
Table 7-3 The porosity ofOPC and HVFA concrete (Water curing) (Unit: % ). 169
Table 7-4 The porosity of OPC and HVF A concrete ( Air curing) ( Unit : % ) ..... 169
Table 7-5 The max. dv/d (log d) and threshold diameter ofOPC and HVFA
concrete (Water curing ) ...................................................................... 174
Table 7-6 The max. dv/d ( log d ) and threshold diameter of OPC and HVF A
concrete ( Air curing) ................................... '" .'" ................................ 174
Table 7-7 The macropore and mesopore ofOPC and HVFA concrete
(Water curing) (Unit: % ) ................................................................. 187
Table 7-8 The macropore and mesopore ofOPC and HVFA concrete
(Air curing) (Unit: % ) ...................................................................... 187
16 2
Table 7-9 High volume fly ash concrete permeability coefficient (Unit: 10. m ). 193
Table 7-10 Influence of air curing on permeability ( Unit: % ) .............................. 196
Table 7-11 High volume fly ash concrete water absorption ( Unit: % ) ................. 203
Table 7-12 Influence of air curing on water absorption (Unit: % ) ....... 205
Table 7-13 X-ray ditrraction ofOPC and HVFA concrete (Water curing)
( 28 days of hydration) ........................................................................ 211
Table 7-14 X-ray ditrraction ofOPC and HVFA concrete (Water curing)
( 90days of hydration ) ......................................................................... 212
Xl
Table 7-15 X-ray dim-action ofHVFA concrete (Water curing)
( 18 months of hydration) .................................................................... 213
Table 7-16 X-ray dim-action ofHVFA concrete (Air curing)
( 18 months of hydration ) .................................................................... 218
xii
List of Figures
xiii
Figure 4-29 The curve for determining the constants a and b (350 kg/m3 mixes ) .... 92
Figure 4-30 The curve for determining the constants a and b ( 450 kg/m3 mixes ) .... 92
Figure 4-3 1 Variation of dynamic modulus of HVF A concrete ( Water curing) ...... 98
Figure 4-32 Variation of dynamic modulus ofHVFA concrete ( Air curing ) ........... 98
Figure 4-33 Effect of total binder content on dynamic modulus (Water curing) ..... 99
Figure 4-34 Effect of total binder content on dynamic modulus ( Air curing ) .......... 99
Figure 4-35 Relationship between dynamic modulus of elasticity and compressive
strength ( Water curing) ... ,.................................... " .......................... 105
Figure 4-36 Relationship between dynamic modulus of elasticity and compressive
strength ( Air curing ) ......................................................................... 105
Figure 4-37 Comparison of predicting modulus of elasticity and published data ..... 108
Figure 4-38 Variation of pulse velocity ofHVFA concrete under water curing ...... 110
Figure 4-39 Variation of pulse velocity ofHVFA concrete under air curing .......... 112
Figure 4-40 Effect of total binder content on pulse velocity ( Water curing) ......... 114
Figure 4-41 Effect of total binder content on pulse velocity ( Air curing) ............. 114
Figure 4-42 Relationship between pulse velocity and dynamic modulus of elasticity
(Water curing) ................................................................................. 117
Figure 4-43 Relationship between pulse velocity and dynamic modulus of elasticity
( Air curing ) ...................................................................................... 118
Figure 4-44 Carbonation depth of concrete - 350 kg/m3 mixes .............................. 121
3
Figure 4-45 Carbonation depth of concrete - 450 kg/m mixes .............................. 123
Figure 4-46 Effect of total binder content on carbonation depth ............................ 123
Figure 4-47 Carbonation of specimens stored in the laboratory - 180 days ............ 124
Figure 4-48 Carbonation of specimens stored in the laboratory - 540 days ............ 124
Figure 4-49 Exposure period versus carbonation depth of concrete -
350 kg/m3 mixes ................................................................................. 127
Figure 4-50 Exposure period versus carbonation depth of concrete-
450 kg/m3 mixes ................................................................................. 127
Figure 4-51 Relationship between carbonation depth and compressive strength ..... 129
Figure 5-1 Comparison of pore classifications ....................................................... 136
Figure 5-2 Relation between the compressive strength of mortar and
gel/space ratio .................................................................................... 139
Figure 5-3 Relation between compressive strength of mortar and of calculated from
the volume of pores larger than 20 run in diameter ................................ 139
Figure 5-4 The composition of the calcium silicate hydrates .................................. 145
Xl\'
Figure 6-1 Photograph of the Micromeritics Poresizer ( Model 9320 ) .................. 152
Figure 6-2 A general view of "The Leeds Permeability Cells" ................................ 157
Figure 6-3 Correction factor related to the length of the specimen ......................... 159
Figure 7-1 Cumulative pore volume curves for the 350 kg/m3 binder content OPC
and HVF A concrete ( Water curing) .................................................... 164
Figure 7-2 Cumulative pore volume curves for the 350 kg/m3 binder content OPC
and HVF A concrete ( Air curing) ........................................................ 165
Figure 7-3 Cumulative pore volume curves for the 450 kg/m3 binder content OPC
and HVFA concrete ( Water curing) .................................................... 166
Figure 7-4 Cumulative pore volume curves for the 450 kg/m3 binder content OPC
and HVF A concrete ( Air curing) ........................................................ 167
Figure 7-5 Differential pore volume curves for the 350 kg/m3 binder content OPC
and HVFA concrete ( Water curing) .................................................... 175
Figure 7-6 Differential pore volume curves for the 350 kg/m3 binder content OPC
and HVF A concrete ( Air curing) ........................................................ 176
Figure 7-7 Differential pore volume curves for the 450 kg/m3 binder content OPC
and HVF A concrete ( Water curing) .................................................... 177
Figure 7-8 Differential pore volume curves for the 450 kg/m3 binder content OPC
and HVF A concrete ( Air curing) ........................................................ 178
Figure 7-9 Cumulative and differential pore volume of 350C/450C, 350F/450F
(Water curing 28 day) ......................................................................... 180
Figure 7-10 Cumulative and differential pore volume of 350SF/450SF, 350S/450S
(Water curing 28 day) ......................................................................... 180
Figure 7-11 Cumulative and differential pore volume 350C/450C, 350F/450F
(Water curing 90 day) ......................................................................... 181
Figure 7-12 Cumulative and differential pore volume of 350SF/450SF, 350S/450S
(Water curing 90 day) .......................................................................... 181
Figure 7-13 Cumulative and differential pore volume of 350F/450F
( Water curing 540 day) ........................................................................ 182
Figure 7-14 Cumulative and differential pore volume of 350SF/450SF, 350S/450S
( Water curing 540 day) ........................................................................ 182
Figure 7-15 Cumulative and differential pore volume of350C/450C, 350F/450F
(Air curing 28 day) ............................................................................... 183
Figure 7-16 Cumulative and differential pore volume of 350SF/450SF, 350S/450S
(Air curing 28 day) ............................................................................... 183
Figure 7-17 Cumulative and differential pore volume of 350C/450C, 350F/450F
(Air curing 90 day) ............................................................................... 184
Figure 7-18 Cumulative and differential pore volume of 350SF/450SF, 350S/450S
( Air curing 90 day) .............................................................................. 184
Figure 7-19 Cumulative and differential pore volume of 350F/450F
( Air curing 540 day) ............................................................................ 185
Figure 7-20 Cumulative and differential pore volume of 350SF/450SF, 350S/450S
(Air curing 540 day) ........................................................................... 185
Figure 7-21 Relationship between compressive strength and porosity ofOPC and
HVFA concrete-Water curing ............................................................ 189
Figure 7-22 Relationship between compressive strength and porosity ofOPC and
HVFA concrete-Air curing ................................................................. 189
Figure 7-23 Relationship between compressive strength and porosity ofOPC and
HVF A concrete .................................................................................. 191
Figure 7-24 Relationship between compressive stren~h and meso-pore ofHVFA
concrete ................................................... " ........................................ 191
Figure 7-25 The permeability coefficient ofHVF A concrete-water curing
( 350 kg/m3 mixes) ............................................................................ 194
Figure 7-26 The permeability coefficient ofHVFA concrete-water curing
( 450 kg/m3 mixes) ............................................................................ 194
Figure 7-27 Relationship between permeability and compressive strength ofHVFA
concrete-water curing ........................................................................ 199
Figure 7-28 Relationship between permeability and compressive strength ofHVFA
concrete-air curing ............................................................................. 199
Figure 7-29 Relationship between permeability and meso-pore ofHVFA concrete
............................................................................................................................. 201
Figure 7-30 The water absorption ofHVF A concrete-water curing
( 350 kg/m3 mixes ) ............................................................................ 204
Figure 7-31 The water absorption ofHVFA concrete-water curing
( 450 kg/m3 mixes) ............................................................................ 204
Figure 7-32 Relationship between permeability and water absorption ofHVF A
concrete ............................................................................................. 207
Figure 7-33 X-ray difITactograms for OPC and HVFA concrete-Water curing
( 28 days hydration) .......................................................................... 214
Figure 7-34 X-ray diffiactograms for OPC and HVFA concrete-Water curing
( 90 days hydration) .......................................................................... 215
Figure 7-35 X-ray diffiactograms for HVF A concrete-Water curing
( 18 months hydration) ...................................................................... 216
Figure 7-36 X-ray diffractograms for HVFA concrete-Air curing
( 18 months hydration) ...................................................................... 219
Figure 8-1 Strategies of using fly ash in HVFA concrete construction ................... 223
NOTATIONS AND ABBREVIATIONS
C=CaO CC = Calcite
S = Si02
C4AH 13 = Tetracalcium Aluminate-13-hydrate
A= Al203 C3 A = Tricalcium Aluminate
F = Fe203 C4AF = Tetracalcium Aluminoferrite
-
C-S-H = C-S-H gel CS = Anhydrite
-
S = S03 Alite = Impure form ofC3S
C = C02 Belite = Impure form of C2 S
-
H=H20 C4AS H 12 = Monosulfoaluminate
-
CH = Calcium Hydroxide C6 AS 3H32 = Ettringite
-
CSH2 = Gypsum AFt = Al203-Fe203-trisulfate
C2ASH8 = Gehlenite Hydrate AFm = Al203-Fe203-monosulfate
B. Abbreviations
"'111
Chapter 1 Introduction 1
CHAPTER ONE
INTRODUCTION
1-l. INTRODUCTION
Pozzolanic materials, such as fly ash ( FA), and cementitious materials such as
ground granulated blast furnace slag ( GGBFS ), are by-products of industry.
Pozzolanic materials have properties similar to those of cement, but are unable to
undergo hydration alone ; thus, they are used as partial replacement of cement, or
blended with portland cement [1,2,3,4]. In a study on synthetic fly ashes, it has been
shown that the final products of fly ash - portland cement reaction are the same as
those of portland cement alone i.e., C-S-H, C4AH13, C8AFH26 and C4ASH12 [5].
There are two major factors which influence the properties of fly ash concrete
- its intrinsic variation in chemical and mineralogical composition, and the philosophy
of mix proportioning adopted to produce the concrete incorporating fly ash [6]. In
most of the studies reported so far, the amount of cement replacement by fly ash is of
the order of 20 % to 30 %. Fly ash is the most abundant of all siliceous materials
available throughout the world. Only a very small proportion of this available fly ash
is used in concrete construction. The aim of this project is to explore the possibilities
of extending the level of cement replacement to much higher amounts so that a much
higher proportion of the available fly ash can be utilised in construction. The
investigation aims to establish the engineering and microstructure properties of this
high volume fly ash ( HVF A ) concrete, compared to that of conventional portland
cement systems. It is hoped that this increased usage of fly ash would help to reduce
disposal problems, and would also contribute, in a minor way, to a cleaner
environment.
Chapter 1 Introduction 2
During the 1950's, construction of both hydroelectric and thermal power stations was
stepped up due to the sudden rise in demand for electricity. The production offly ash
as an industrial by-product from these power generating stations also gave rise to
studies connected with the use of fly ash as a cement replacement material. The
utilisation of fly ash in European countries began at about the 1960's, particularly in
the US SR and Great Britain.
The advantages of incorporating a good quality fly ash in concrete include the
improvement of the workability of the concrete, improvement of strength, water
tightness and durability at later ages, reduction in the heat of hydration, and reduction
of drying shrinkage. Probably, the foremost advantage of fly ash lies in the economy
resulting from reduction of cement content and the production of better quality
concrete. As shown in published literature, these advantages of fly ash compensate
for much of the defects of portland cement, and the use of good quality fly ash as an
essential ingredient of structural concrete is now widely accepted. From the point of
view of using relatively large volumes of fly ash in portland cement concrete, this
concept has many beneficial implications in terms of usage of waste materials,
reduction of environmental pollution, and the conservation of material resources.
As mentioned before, much of the earlier research on fly ash concrete concentrated
on the development of concretes with about 20 ~ 30 % cement replacement. More
recently, in the last couple of decades, a limited amount of research has been carried
out on concretes incorporating higher amounts of fly ash, from 50 % to 60 % by
mass of portland cement [7]. However, there is the need for much more information
on such HVF A concrete to establish their properties reliably, and to give confidence
to suppliers and users. In the present study, the proportion of fly ash in the total
2
Chapter 1 Introduction 3
The investigation is divided into two parts. Part I of the research deals with
the basic characteristics of concrete containing high volume fly ash. The concrete was
designed using different cementitious contents and replacement methods. In order to
investigate changes in engineering properties, the concretes were cured under two
different conditions, namely, prolonged water curing and prolonged air drying after
an initial 7 day water curing. Silica fume was also incorporated in some of the mixes,
and fly ash was used to replace both cement and fine aggregate. A superplastiscizer
was also incorporated to enhance workability, and to keep the water content low. In
addition to the strength development of high volume fly ash concrete, several other
concrete properties were also investigated. To complement the studies of engineering
properties, Part II of the project concentrates on the hydration reaction and
microstructure development of high volume fly ash concrete. Changes in pore
structure and mineralogical composition were also investigated along with
permeability, water absorption and the hydrate products in the pozzolanic reaction.
The main aim and objective of this project is to investigate the engineering
properties, microstructural characteristics and the hydrate products of concrete
containing high volume fly ash. The individual aims and objectives of each part are
given below :
2. To study the short and long term strength development under different curing
conditions and mixture parameters of high volume fly ash concrete.
3
Chapter 1 Introduction
4. To evaluate the effectiveness of moist and dry curing regimes on expansion and
drying shrinkage behaviour ofHVF A concrete.
2. To study the pore size distribution and porosity in relation to the engineering
3. To seek out the changes in calcium silicate hydrate ( C-S-H gel ), with curing
ages in the hydration reaction of hardened HVF A concrete.
In the third chapter, the experimental programme deals with the details of test
parameters, concrete mixture design, specimen preparation and the experimental
procedure. The characteristics and properties of the raw materials used in the
experimental part of the research such as cement, fly ash, aggregate, superplasticizer
and silica fume are also reported.
In chapter four, the experimental results of tests from Part I relating to the
laboratory investigation of the engineering properties of high volume fly ash concrete
are discussed. Long and short term strength development along with the dynamic
modulus of elasticity and ultrasonic pulse velocity of different types of concrete are
investigated. Other high volume fly ash concrete engineering properties such as
swelling, and drying shrinkage are also presented. The effects of carbon dioxide from
the atmosphere on the depth of carbonation in high volume fly ash concrete is also
discussed in this section. The relationship between strength and other properties and
cementitious content for different replacement methods is also presented.
Part II of the project is covered in chapters five, six, and seven. In chapter
five, the review deals with a brief analysis of the pore structure of hydrated cement
paste. This is followed by a detailed review of the influence of porosity and pore size
distribution on the engineering properties. A detailed review of the hydrate products
of hardened cement paste is also presented.
Chapter six deals with the test details, the preparation of test specimens and
the tests conducted on the hydration and microstructure of high volume fly ash
5
,~----~
Chapter J Introduction 6
concrete. The principal tests used included mercury intrusion porosimetry, aIr
permeability, water absorption test and X-ray diffraction.
Chapter seven describes the results from the tests of Part II. This section
looks into the microstructure mechanism in order to confirm the results from the
engineering properties described in Chapter four of the thesis. This also contains a
detailed discussion describing the changes in the mineralogical phase of different
cementitious contents during the early and later periods of hydration of high volume
fly ash concrete. It also presents a discussion on the mechanisms of pore size
distribution.
Chapter eight summanzes the overall conclusions obtained from the two
experimental projects along with some recommendations for future work towards the
commercialisation of high volume fly ash concrete systems.
6
Chapfe-rLITterature Review
CHAPTER TWO
LITERATURE REVIEW
2-1. INTRODUCTION
The utilisation of fly ash in European countries came at about the 1960's. In
the USSR and Great Britain, standards on fly ash were approved in 1963 and 1965
respectively. The first Australian Standard for fly ash in concrete was published in
1971. This was in the form of specified requirements of fly ash, and guidance for the
use of fly ash in concrete. The Canada Centre for Mineral and Energy Technology (
CANMET ) began work in 1985 to develop structural concrete incorporating high
7
Chapter 2 Literature Review 8
volumes of low calcium fly ash. In these investigations, an ASTM Class F fly ash was
used in proportions of 50 to 60 percent of total cementitious material [7]. Low unit
water content and a high degree of workability were obtained using high range water
reducing admixtures. Full details of the study of high volume fly ash concrete were
presented by EPRI ( Electric Power Research Institute) in 1993 [9].
The advantages of good quality fly ash are the improvement of workability of
concrete, improvement of strength, water tightness and durability of concrete at
advanced ages, alleviation of heat of hardening of concrete and reduction of drying
shrinkage, although probably, the foremost advantage of fly ash lies in the economy
resulting from the reduction of cement. When considering the fact that thermal power
stations have to incur great expense in disposing of the material as waste, the
advantages of utilising fly ash are even more marked.
8
Chapter"2 Literature Review 9
The chemical composition of fly ash is determined by the coal used. More than 85
percent of most fly ashes comprise compounds and glasses formed from Si02,
A1203, Fe203, CaO and MgO. Kokubu [11] had tested the chemical composition of
fly ash in various countries as shown in Table 2-1, which emphasizes the many
differences in fly ash composition according to the country of production. The
average values of the important oxides are 40 ~ 58 % for Si02, 21 ~ 27 % for
A1203, 4 ~ 17 % for Fe203 and 4 ~ 6 % for CaO. These variations are largely due to
the differences in the quality of coal used in the various countries.
Physical characteristics
In general, fly ash consists of glassy spheres of sizes varying from under 1 Il m
to as large as 150 Ilm, although typical particle size distribution shows that most of
the material is under 20 Ilm [12]. The range of particle sizes in any given fly ash is
largely determined by the dust collection equipment used. Table 2-1 gives the specific
surface areas of fly ash in various countries. In this case also, a wide difference is
seen between countries, the value ranging from 1220 to 8100 cm2/g.
1. Most of the fly ash particles occur as solid glass spheres, which are mostly
transparent but may be partially devitrified.
2. Irregular masses of large size may exist either as agglomerates of small silicate
glass spheres or porous particles of incompletely burnt carbonaceous matter.
9
Chapter 2 Literature Review 10
4. Sometimes low-calcium fly ashes may contain a small amount of hollow particles
which are either completely empty ( cenospheres ) or packed inside with smaller
spheres ( plerospheres ) [12].
Country Number oj Loss Si02 Al203 Fe203 CaO MgO S03 Na20 K20 Specific
Composite on Surface
Samples Ignitiol1 (cm2 /g
Japan 12
Av. 0.73 57.96 25.86 4.31 3.98 1.58 0.34 1.49 2.15 3090
Max. 1.23 63.27 28.35 5.90 6.74 2.09 0.81 2.36 3.15 4150
Min. 0.06 53.41 22.88 2.82 1.04 1.00 0.02 0.88 1.73 1220
USA 34
Av. 7.83 44.11 20.81 17.49 4.75 1.12 1.19 0.73 1.97 3673
Max. 18.00 51.90 28.30 31.30 12.0 1.40 2.80 2.10 2.98 4795
Min. 1.00 32.70 14.60 8.50 11.1 0.70 0.30 0.22 1.28 2430
UK 14
Av. 3.86 46.16 26.99 10.44 3.06 1.96 1.59 0.90 3.26 5180
Max. 11.70 50.70 34.10 13.50 7.70 2.90 6.80 1.90 4.20 8100
Min. 0.60 41.40 23.90 6.40 1.70 1.40 0.60 0.20 1.80 2500
France 17
Av. 3.72 48.45 25.89 8.07 5.95 2.36 1.01 0.64 3.94 ---
Max. 15.15 54.05 33.40 15.30 38.7 4.45 7.00 0.85 6.00 ---
Min. 0.30 29.90 10.80 5.80 1.48 1.10 0.10 0.15 0.70 ---
Germany 9
Av. 9.65 41.13 24.39 13.93 5.06 1.85 0.77 --- --- ---
Max. 20.10 49.54 29.35 20.88 11.8 4.26 2.10 --- --- ---
Min. 1.48 34.10 21.06 8.37 2.18 0.75 0.12 --- --- ---
USSR 15
Av. --- 55.08 25.97 7.83 5.08 1.81 1.63 --- --- ---
Max. --- 62.08 37.15 12.01 10.6 2.90 3.78 --- --- ---
Min. --- 47.90 20.71 3.08 1.10 0.28 0.20 --- --- ---
10
Chapier 2 Literature Review II
The low-calcium fly ashes, due to the high proportions of silica and alumina,
consist principally of alumino silicate glass. In the furnace very large spheres of
molten glass may not get cooled rapidly and uniformly, thus crystallisation of
aluminosilicates such as silica-alumina ( Al203 • Si02 ) and mullite ( 3Al203 • 2Si02
) as slender needles, takes place in the interior of the glassy sphere. This partial
devitrification of glass in low-lime fly ashes, therefore, would account for the
presence of crystalline aluminosilicates which are relatively non-reactive. In regard to
high calcium fly ashes which may also contain significant amounts of magnesia,
alkalis and sulphates, the situation is more complex than the low-calcium ashes from
anthracite or bituminous coal. The chemical composition and reactivity of the non-
crystalline or glassy phase is different from the glassy phase typically present in low-
calcium fly ashes. Also, the principal crystalline mineral in high-calcium fly ash is
generally C3A, which is the most reactive mineral present in portland cement.
11
Chapter 2 LiteratureReview 12
12
Chapter 2 Literature Review 13
But in general the C / S ratio ofC-S-H gel might be in the range between 1.0
~ 2.0 with an average of 1.5. According to Young [13] the pore structure of a
saturated paste is more open than previously thought and that the microporosity may
only be created during drying.
( a ) Tricalcium aluminate
The primary initial reaction of C3A with water in the presence of a plentiful
supply of gypsum is :
13
Chapter 2 Literature Review
- -
C3 A + 3 CS H2 + 26 H ~ C3 A • 3 CS • H32 ------------- ( 2-3 )
- -
2 C3A + C3A. 3 CS • H32 + 4 H ~ 3 • C3A. CS • H12 -------- (2-4 )
- -
C3A + CSH2 + 10 H ~ C4ASH12 ---------------------- (2-5 )
The monosulphate is a new source of sulphate ions from the solution, the
ettrignite can be formed once again :
The more gypsum there is in the system, the longer the ettringite will remain
stable.
l-l
Chapter 2 Literature Review 15
(b ) Tetra - calcium alumino-ferrites (Ferrite phase) --- In this case, the AIF
ratio is not the same as that of the parent compound.
In the presence of gypsum, the reactions are slower and evolve less heat. The
hydration between C4AF and sulphate ions may be represented as :
- -
3 C4AF + 12 CSH2 + 110 H ~ 4 [ C6 (A,F) S 3H32 ] + 2 (A,F ) H3 ----- (2-9 )
- -
3 C4AF + 2 [C6 (A,F) S 3H32] + 14 H ~ 6 [C4 (A,F) SH12] +2 (A,F) H3
----- ( 2-10 )
The above equation such as ( A,F ) means that A1203 and Fe203 occur
interchangeable in the compound.
15
Chapter 2 literature Review 16
The volume of the hydration products in a full hydrated and saturated paste is
listed in Table 2-2.
Pozzolans are rich in Si02 and poor in CaO, whereas the opposite holds for
clinker and portland cement. Calcium hydroxide is produced from the excess CaO
16
Chapter 2 Literature Review 17
during the hydration of cement constituents, and serves as the mam source of
alkalinity for the pozzolanic activity, i.e., for the conversion of the Si02 of the
pozzolan into C-S-H. The pozzolanic reactions of various pozzolans have been
examined in several investigations previously [11,14,15,16]. Guillaume [1 7] found a
correlation between the decrease in calcium hydroxide, produced by hydration of
cement, and the extent of pozzolanic reaction, which suggested that the pozzolanic
reaction of fly ash starts at about 14 days. Bache et a1 [18] stated further that when
fly ash absorbs calcium hydroxide, the absorbed calcium hydroxide appears to form a
membrane over the fly ash particles and a film of water exists at the interface between
the membrane and fly ash particles.
A study on synthetic fly ashes suggested that the initial reaction products
appeared to be poorly-crystallized hydrates of calcium silicate or calcium aluminate.
After one week of curing, a tetracalcium aluminate hydrate ( C4AHx ) was identified
along with some fine needle-like crystals which were probably ettringite. After eight
weeks, C-S-H was identified. The final products of fly ash reaction are the same
those of portland cement. So, in very simplified terms, pozzolanic activity can be
descried by the following reactions [5][19] :
It is worth noticing that the reaction of fly ash in portland cement coincides
with the reaction of portland cement. The main difference between the portland
cement hydration reaction and the pozzolanic reaction lies in the former, where the
entire quantity of each basic oxide of portland cement, i. e., of C, S, A and F, reacts
in the form of the compounds, C3S, C2 S, C3A, and C4 AF , in which they occur in the
clinker, whereas in the latter, only the amorphous forms of these pozzolanic oxides
17
Chapter 2 Literature Review 18
may react with lime solution ( CH ). Another difference is that the pozzolanic
reactions of fly ash may be initiated by the adsorption of calcium hydroxide, produced
by cement hydration, on the surfaces of the fly ash particles. The adsorbed calcium
hydroxide appears to form a membrane over the fly ash particles and a film of water
exists at the interface between the membrane and fly ash particles [20].
18
Chapter 2 Literature Review 19
Hydration occurs right after the cementitious materials come into contact with water.
The hydraulic cementitious reaction progresses to make the cementitious system
harden and develop strength. Basically, setting behaviour is the result of cementitious
grains dispersing and hydrating with water, and gradually forming a solid / liquid
suspension of various size hydrates. As cement hydrates, the inner structure of the
cementitious paste is further reinforced, and becomes a network structure which
makes the cementitious material set and gain strength.
Bleeding is mainly due to the amount of mix water being greater than that
required for complete cement hydration. The phenomenon is caused by the solid /
liquid suspension being generated after the mixing of cement and water. Except for
the adsorption of water due to the electrical double layer and diffusive water layer,
the extra water escapes from the structural network of cementitious pastes, which
explains the mechanism of bleeding. Concrete using fly ash is generally reported to
show reduced segregation and bleeding, and to be more satisfactory when placed by
pumping than plain concrete placed under the same circumstances.
The pozzolanic activity of fly ash has a close relationship with its fineness. The small
size and the essentially spherical form of the particles comprising fly ash usually
influence the rheological properties of cementitious pastes, causing a reduction in the
amount of water required for a given degree of workability from that required for an
equivalent paste without fly ash [24]. For a constant workability, the reduction in the
water demand of concrete due to fly ash is usually between 5 and 15 % by
19
Chapler 2 Literature Review 20
comparison with a portland cement only mix having the same cementitious material
content. There are many reports confirming the workability and fineness
characteristics offly ash. As early as 1937, Davis et al made this clear in their paper
[8]. Kokubu and Yamada [19] showed that unit water-content ratio was lowered
with increasing fineness and that the unit water-content ratio exceeds 100 percent
when specific surface area was lower than 2000 cm2/g with almost no effect on
workability. Meanwhile, they cited the results from Homms et aI, who collected 14
varieties of fly ash varying in fineness between 1000 ~ 4150 cm2/g from a thermal
power plant and performed tests on mortar holding cement-fly ash quantities
constant, replacing 25 percent of the cement with fly ash and determining unit water
contents to obtain equal consistency. Cabrera and co-workers [25] have reported on
measuring the effects of seven fly ash on concrete workability by using Tattersall
Two Point Test. The test programme shows that fly ashes vary in their ability to
improve concrete workability. There was no question that the fineness values of a fly
ash affected the workability of the concrete. However, the standard value of fineness
is significant in providing a criterion in planning fly ash collecting facilities. Table 2-1
also gives examples of the specific surface areas of fly ash in various countries.
20
Chapter 2 Literature Review 21
C2 S components of portland cement takes place more slowly than C3S reaction and
the approximate reaction rate ofC3S, This combined with the dilution of the portland
cement component by fly ash results in a reduced rate of heat evolution, and reduced
ultimate heat of hydration. Sturrup et al [26] used a conduction calorimeter, and
found that fly ash could be used successfully in lieu of both ASTM Type II and IV
moderate and low-heat cements to control temperature rise and thermally induced
cracking in mass concrete. The use of fly ash to reduce the heat of hydration of
concrete is well established. The US Bureau of Reclamation ( USBR ) has conducted
numerous experiments concerning heat of hardening of concrete containing fly ash.
Malhotra et al also cited the results from Eifert who showed the effect of fly ash and
a calcined diatomaceous shale on the temperature rise of mass concrete, and found
that for a given total cementitious materials content, high volume fly ash concrete
exhibits a substantially lower temperature rise than conventional concrete [24,27].
The factors affecting the development of concrete cementitious strength include its
internal water content, curing condition, mix proportioning and admixtures. The
particles of fly ash in a dry state often present a floc structure from the mutual
attraction of the particles ; when concrete is mixed using fly ash which has a floc
structure, or present as agglomerates, the particles of fly ash are flocculated and are
difficult to disperse. Meanwhile, the contribution which fly ash makes to strength
development of concrete is the result of pozzolanic reaction between the soluble
silica and alumina from the glass phase of the fly ash, and lime, to form calcium
silicate and calcium aluminate hydrates. Moreover, in addition to the effect of
chemical reaction, fly ash has a physical effect of improving the microstructure of the
hydrated cement paste. The main physical action is that of packing of the fly ash
21
Chapter 2 Literature Review 22
particles at the interface of coarse aggregate particles. Much research has shown that
proper percentage replacement of portland cement in concrete results in comparable
strength between normal OPC and fly ash concrete [12,14,21,24,28]. In
consideration of early strength fly ash concrete, it has been reported that concrete
mixes with normal weight and lightweight aggregates containing 30 percent by
weight of cement replacement of fly ash can be designed to have adequate early one-
day strength for structure applications. Swamy et al [21] and Cripwell et al [29] have
reported on compressive strengths when a part of normal portland cement is replaced
with fly ash ; these studies show that the later-age strength development of fly ash
concrete is higher than that of normal OPC concrete resulting in an 18 percent
increase in compressive strength at 300 days. Thomas and Matthews [30] also have
reported on investigating existing concrete structures built with fly ash concrete. The
age of structures ranged from 2 to 33 years. The results indicated that in all cases the
fly ash concrete has greater long term strength by comparison with the OPC concrete
in the same structure. These effects of sand replacement on early age and long term
fly ash concrete strength also suggests that the superior performance of these
mixtures compared to normal OPC concrete mixtures is attributable to the
densification of the paste structure due to pozzolanic action between the fly ash and
calcium hydroxide liberated as a result of hydration of cement [31].
In recent years there has been considerable interest in the production and use
of very high strength concrete ( 50 ~ 60 MPa ) at 28 days. This type of concrete is a
particular example of the combined practical use of fly ash and water-reducing
admixtures [24]. An extensive scientific and engineering study ( 1990 - 1993 ) by
EPRI ( Electric Power Research Institute) has demonstrated the use of high volume
fly ash ( HVF A ) concrete produced for structural applications. High volume fly ash
concrete has adequate early age strength, and very good later age strength ; and, the
ratio of the flexural to compressive strengths, and of splitting-tensile to compressive
strength are comparable to the values for conventional portland cement concrete [9].
22
Chapter 2 Literature Review 23
The volume of a cementitious paste varies with its water content. Drying will cause a
volume decrease, i. e., drying shrinkage ( or shrinkage ), and wetting causes volume
increase, i. e., swelling. Consequently, the rate of removal of water from the
cementitious paste controls the rate of shrinkage of the cementitious paste. All
factors which affect drying, such as temperature, humidity, pressure and arr
movement, will also influence shrinkage. Also, the factors that affect the physical and
chemical properties of the paste, such as cementitious composition, admixtures,
curing temperature, and water / cementitious ratio, will also affect shrinkage.
2-6-3. Creep
Some studies have shown that fly ash increases the creep of concrete [28]. However,
these investigations involved direct replacement of fly ash for cement, and produced
lower strength at loading. Since creep is influenced by compressive strength and the
modulus of elasticity of the concrete, higher creep strains were found in fly ash
concrete at early age loading when strength was low, however the rate of creep
decreased at later ages.
When similar 28 day strength is compared, creep of fly ash concrete is lower
than its plain concrete equivalent. Investigations by Bamforth [32] on mass concrete
showed that a reduction of 50 % in creep can be obtained when cement is replaced by
about 30 % fly ash. Ross [33] had obtained 15 % reduction with 25 % fly ash
replacement.
The high volume fly ash concrete appears to produce low creep strains
compared to those of normal concretes. The low creep strains could be the result of
a large portion of fly ash remaining unreacted in the concrete, and thus acting as fine
aggregate, providing increased restraint against creep. The long-term creep behaviour
of concrete containing high volume fly ash and superplasticizer has been described by
23
Chapter 2 Literature Review
Cripwell et al [29]. The combined effect of high volume fly ash and superplasticizer is
to decrease the creep of concrete stored in water or stored in air; after 300 days of
loading, creep was 50 % less than that in the case of portland cement concrete.
Similar, long-term creep results by Alexander et al [34], showed that the creep of
plain concrete specimens loaded at stress / strength ratio of 0.25 and 0.4 were found
to be 50 % greater than that of fly ash concrete at 2 years, and the situation remained
unchanged at 6 years.
Modulus of elasticity is very important from a design point of view in reinforced and
prestressed concrete, since it can control concrete behaviour as much as compressive
strength, particularly in structural elements subjected to flexure. The modulus of
elasticity of concrete follows a pattern similar to compressive strength. Fly ash
concrete is known to exhibit similar or slightly higher modulus of elasticity than
portland cement concrete [12,24,26,35,36]. Lorig term, higher strength concrete can
help to reduce deflections under load, and these can be anticipated when fly ash is
used in concrete. A report by EPRI [9] on high volume fly ash concrete claims that,
the generally high modulus of elasticity values of this concrete seems to be due to the
densifying effect of the unhydrated fly ash particles acting as fine aggregate.
2-7-1. Permeability
Factors affecting permeability amongst others, include the presence of large voids in
the hydrated cement, microcracks at the aggregate-cementitious paste transition
24
Chapter 2 Literature Review
zone, water / cementitious ratio and curing efficiency. Unless curing conditions are
especially poor, the impermeability of concrete containing fly ash is improved. The
pozzolanic reaction of fly ash in concrete, which produces calcium silicate hydrate,
tends to fill these unoccupied spaces to form a product with decreased permeability,
which also results in improved corrosion resistance.
25
Chapter 2 Literature Review 26
Most commonly, sulphate attack is a phenomenon in which the sulphates from the
environment surrounding the concrete react with reactive calcium aluminate phases
within the hardened concrete, resulting in internal expansion due to the formation of
the crystalline compound ettringite, and yields a greater volume causing
deterioration. The replacement of a portion of the portland cement with fly ash on the
sulphate resistance of concrete can be conveniently explained in terms of two
mechanisms. First, the pozzolanic reaction reduces the amount of calcium sulphate,
which consequently decreases the amount of gypsum formed. The second
mechanism, the concrete with fly ash in the mixture can decrease the proportion of
the larger pores which results in a less porous matrix, and decreases the total amount
oftricalcium aluminate (C3A) content of the cement.
In recent years, several studies have tried to document the effect of fly ash on
the sulphate resistance of concrete [35]. Mehta [38] and Carrasquillo et al [35]
indicate that the effect of fly ash on the sulphate resistance of concrete is related to
the composition of both the crystalline and amorphous phases. This phase dependent
interaction can be used to expand on the R-value concept to reliably predict the effect
of fly ash on the sulphate resistance of concrete. USBR ( United States Bureau of
Reclamation) guidelines include a resistance factor (R-value ) defined as :
CaO%-5
R= ---------- ( 2 - 18 )
Fe203%
and found that using cement replacement levels of 25 % fly ash with R-values less
than 1.5 was found to improve the sulphate resistance of the concrete, and suggested
that with the calcium oxide ( CaO ) content less than 5 % and low ferric oxide (
Fe203 ) contents in the concrete, sulphate attack is reduced. Tikalsky and
Carrasquillo [35,39] have given the same conclusion. In 1993, EPRI [9] reported the
results of sulphate resistance by using 12 different power plant fly ashes on high
volume fly ash concrete studies. From this work it was concluded that all of the fly
ashes tested greatly improved sulphate resistance.
26
Chapter 2 Literature Review 27
This undesirable expansion of concrete is due to reaction between the aggregate and
hydroxyl and alkali ions in the pore solution forming a gelatinous alkali silicate, which
is capable of absorbing large quantities of water. The consequent expansion disrupts
the concrete and reduces both the strength and elasticity properties of concrete and
the structural integrity of concrete structures [40,41,42]. Fly ash with siliceous
constituents may react with alkalis in the cement to form products which take up
water and swell, and can minimize or prevent the expansion [28]. The use of
pozzolans to prevent alkali-aggregate reaction has been summarized from the results
of Pepper and Mather by Plowman and Harrogate. They concluded that the
effectiveness of different fly ashes in reducing long-term expansion varied widely, and
suggested that the effectiveness may be dependent upon their alkali content or
fineness [40].
inhibiting the expansion due to alkali-aggregate reaction. The rationale for reduction
in the expansion due to alkali-aggregate reaction is the extremely low permeability of
the high-volume systems, and the reduced pH of the pore solution. Similar results by
Alasali & Malhotra [45] showed that the expansion of the concrete incorporating 58
% fly ash, after 275 days of exposure under the various test conditions, is
insignificant.
While it is clear that some fly ashes are effective in controlling alkali-
aggregate expansion, it is questionable whether the strength losses caused by
replacement of the cement by fly ash would be tolerable for more than a limited
number of applications. The use of such large amounts of fly ash to reduce alkali-
aggregate interaction, if required, would certainly demand reproportioning of
concrete if acceptable early strength were to be attained.
27
Chapter 2 Literature Review 28
2-7-4. Carbonation
Carbonation is the process whereby carbon dioxide from the atmosphere penetrates
the concrete, converting the hydroxides to carbonates and lowering the alkalinity of
the protective concrete surrounding the steel, leading to dissolution of the protective
oxide layer on the steel bars and to reinforcement corrosion [5,45]. There are a large
number of factors that affect the rate and extent of carbonation, such as, the
concentration of Ca(OH)2 produced by the hydration of portland cement as well as
C-S-H, which contains CaO. With concrete that contains portland cement and fly ash,
there is a further increase in the number of carbonating variables when some of the
cement is replaced by fly ash [46]. Browne [47] cited the results from Mattews,
obtained at the Building Research Establishment, who showed that after 10 years, fly
ash concrete carbonates approximately to the same extent as normal OPC concrete,
with the rate of carbonation found to be related to the strength of the concrete.
Langleyet al [27] carried out tests on large concrete blocks containing high volumes
of low-calcium fly ash and showed that, for concretes of similar 730 days strength, fly
ash concrete carbonates to a lesser depth than the OPC concrete. After 7.5 years
cores of high volume fly ash concrete examined showed average carbonation depth
less than 8 mm [44].
The main cause of reinforcement corrosion is the loss of efficiency of the concrete
cover, which protects the steel and keeps it in a passive state. Chloride ions have the
28
Chapter 2 Literature Review 29
special ability to destroy the passive oxide film of steel. The chloride ions which
penetrate into the concrete arise either from internal mix constituents or external
sources. Internal mix constituents include CaCl2 added as an accelerating admixture
and other admixture compositions. External sources are contact with sea water, de-
icing salts and structures exposed to the splash zone and marine environment.
The paths that allow chloride ions to intrude depend upon the pore structure
of the concrete. The intrusion process of chloride ions occurs by external hydraulic
pressure that bring the ions through the pores into concrete. The amount of chloride
required to initiate corrosion depends on the pH of the solution in contact with the
steel, and comparatively small quantities are needed to offset the basicity of portland
cement [13]. Even at high alkalinities, the intrusion of chloride will be accompanied
by a chemical reaction with the tricalcium aluminate to form calcium-chloroaluminate
hydrate. There are several strategies that have been developed to prevent corrosion
of reinforcing steel, based largely on studies carried out on bridge-decks. The
possible strategies can be divided into several categories such as reduction of
permeability of concrete, protective coatings on the concrete, protective coatings on
the steel, and suppression of the electrochemical process. Swamy and Tanikawa [49]
have given recommendations for the selection criteria of surface coatings, as shown
in Table 2-3. The effect of using fly ash in concrete is to reduce the amount of
chloride intrusion which can arise due to a lower porosity and high tortuosity of the
pore channels inhibiting ionic diffusion. Dhir et al [50] found that it is not the fly ash
quality that effects chloride diffusion but the quantity of fly ash in the concrete.
Haque, Kayyali, and Gopalan [51] also observed that the presence of fly ash
significantly reduced the concentration of free CL- ions in concrete when the chloride
ingress is from the surrounding sodium chloride solution. The rapid chloride-ion
diffusion tests performed on HVFA concrete indicated very low chloride-ion
diffusion at 1 year, ranging from 119 to 179 coulombs confirming the excellent
resistance ofHVFA concrete to chloride-ion penetration [9].
29
Chapter 2 Literature Review 30
Silica fume is the by-product of electric arc furnaces during the production of silicon
metal or ferrosilicon alloys. Ferrosilicon alloys containing 50 ~ 70 % Si are the most
. important alloys of silicon. The raw materials used for the production of ferrosilicon
are usually 3 '" 6 inch pieces of crushed quartz, coke, and iron ore pure silicon metal
is produced at temperatures of the order of 2000°C, in the areas between electrodes
submerged in the charge. About 10 '" 15 % quartz in the charge is lost in the form of
Si and SiO vapours, which react with air to form Si02· The fine particles are
collected by filtering gases escaping from the arc furnace. The chemical compositions
of the condensed silica fume from the ferrosilicon industry corresponds, generally, to
85 '" 92 % Si02, 0.5 '" 3 % Fe203, 1 ~ 3 % alkalis, 1 ~ 2 % carbon, and minor
30
Chapter 2 Literature Review 31
amounts of Al203, CaO, and MgO [52]. These percentage compositions depend on
what is being produced in the furnace. The particle size analyses show individual
spherical particle shape with a number of primary agglomerates, and amorphous
particles range from 0.01 to 0.3 ~m, with about 70 % < 0.1 ~m. Nitrogen
adsorption measurements show 15 - 20 m2/g surface area of the material.
Areas of application where silica fume has shown particular advantage, and
some of the results from these investigations have been recently published in a
comprehensive review [15,43,53,54,55,56]. There is strong evidence that the use of
silica fume increases the homogeneity and decreases the number of large pores in
cement paste, and results in a denser interface between cement paste and coarse
aggregate. These studies have demonstrated that the pozzolanic activity effect of
silica fume in concrete is faster than ordinary portland cement, because of its high
surface area, and the high amount content of amorphous silica in silica fume. Many
authors foresee a wide-ranging application of silica fume if it is used as a strength
accelerator in portland cement concrete containing fly ash [52].
The factors affecting the development of cement paste strength include its internal
water content, voids in the matrix, and interface organisation. Reducing water
content and voids as well as compressing the interface may produce high strength /
high performance concrete. Concrete is commonly used in constructing building
frames. Recent developments in high strength / high performance concrete have
extended its use to tall buildings [57].
The name "high strength / high performance concrete" implies that, in many
cases, it is high durability that is the required property, although, in others, it is high
strength, either very early, or at 28 days, or even later. In some applications, a high
modulus of elasticity is the property sought.
31
Chapter 2 Literature Review 32
Although both properties of high strength and low permeability are not
necessarily concomitant, high strength requires a low volume of pores, especially of
the large capillary pores. The way to have a low volume of pores is, in general,
achieved by the use of mineral admixtures, such as fly ash, slag or silica fume, which
fills the space between the cement particles and between the aggregate and the
cement particles.
\
(
Table 2-4. Typical high compressive strength of fly ash concrete [58J
(Unit: MPa)
32
Chapter 2 Literature Review 33
Concrete containing high volume low-calcium fly ash ( ASTM Class F ) has
been used in various parts of the world during the last 10 years. The first use of this
type of HVF A concrete is believed to have been in roller compacted dams and
highway base courses. Roller compacted concrete has very low unit water content,
and lacks the workability required for placement as structural concrete.
Swamy and Mahmud [59] also introduced such HVFA concrete at the
international symposium on the Fly Ash, Slag, Silica Fume and Natural Pozzolans in
Concrete of 1986. Subsequently, W. S. Langley and Association limited, Halifax,
Nova Sotia undertook in a large HVFA concrete block, 3.05x3.05x3.05 m, to
monitor the long-term strength development of HVF A concrete. The water-to-
cementitious material ratio of these HVF A concrete ranged from 0.27 to 0.49, and
3
their total cementitious material content ranged from 400 to 225 kg/m . Test results
were presented in 1989 and 1992 respectively [27,60]. The test data on strength
33
Chapter 2 Literature Review 3-1
investigated. The creep and drying shrinkage of the HVF A concretes investigated
were relatively low.
35
Chapter 2 Literature Review
36
Table 2-5 Mix proportions and engineering properties of some HVFA concretes
Flexural strength 100 x 100 100 x 100 100 x 100 r6x102x40 150x30G 150x30C 150x30C 150x30C 150x300
(MPa) x500mm x500mm x500mm 6mm mm mm mm nun mm
at age ( days) pnsms prisms pnsms lCylinders o..,ylinders (;ylinders lCylindeTI cylinders
1 0.78 1.92 3.12 - -- - -- -- --
7 1.80 3.12 3.92 - - - -- - -
Modulus of e1asticit) 100xlOO 100 x 100 100 x 100 152x305m 150x30G 150x30C 150x30C 150x30C 150x300
(GPa) x500mm x500mm x500mm mn mm mm mm mm mm
at age ( days ) pnsms prisms pnsms cylinders lCylinders lCylindeTI lCylinders ~ylinders cylinders
28 35.6 41.7 46.9 35.5 36.1 35.1 33.8 27.9 -
36
Chapter 2 Literature Review 37
37
Chapter 3 Experimental Programme 38
CHAPTER THREE
EXPERIMENTAL PROGRAMME
3-1. INTRODUCTION
In order to understand the role of fly ash in concrete, and to develop a good method
of mixture proportioning of high volume fly ash concrete ( HVF A ), several mixtures
were tested and the results analysed. The development of strength in high volume fly
ash concrete is intimately related to mix proportioning, and the method of mix
proportioning will influence the properties of such concrete. The underlying principle
of the method used here is to produce high performance concrete for structural
applications. The first stage of this research consists in the development of several
mixtures of high volume fly ash concrete and the details are presented below.
,I
Several methods of mixture proportioning of fly ash concrete have been developed
over the years. Basically, fly ash has been used as partial replacement of cement, fine
aggregate, or cement and fine aggregate. This method of mixture design has been
used extensively in many studies, and is also adopted for use in this research. The mix
parameters and proportions of the cement replacement mixtures are summarized in
Table 3-1 and Table 3-2.
38
Chapter 3 Experimental Programme 39
water reducer ( fIR WR ) was used in all the mixes, and the dosage of this
superplasticizer was adjusted to give a slump of 150 '" 200 mm.
3
Table 3-2 High volume fly ash concrete mixtures (Unit: kg/m )
ash.
39
Chapter 3 Experimental Programme 40
Cement
The cement used was ordinary portland cement ( oPc ) conforming to British
Standard obtained from Castle Cement Limited, Lancashire. The specific surface area
of this cement was 345 m2/kg.
Fly ash
A Class F fly ash with the ASTM designation C618 was used in this
investigation. The fly ash used in this concrete was obtained from Boral Pozzolan
Ltd, Herts. The MgO content of l.62%, satisfies to the BS 3892 : Part 1 : 1981
"Pulverized-fuel ash for use as a cementitious component in structural concrete",
which restricts the MgO content to less than 4 %.
Silica fume
A silica fume slurry containing 50 % by weight of water was used. The silica
fume was supplied by Elkem Materials Limited in slurry form. Its specific gravity is
2.2. The chemical composition of the portland cement and mineral admixtures are
shown in Table 3-3.
Normal Aggregate
The fine and coarse aggregates were obtained from the Redlands quarry. A
sieve analysis was carried out on representative samples in accordance with BS 812 :
Part 103. 1: 1985 "Sieve tests". The sieve analysis results of the aggregates are given
.to
Chapter 3 Experimental Programme
in Table 3-4. The fine aggregate had a water absorption coefficient of 2.1 %, and a
bulk specific gravity 2.62, in saturated surface dry condition. The coarse aggregate
consisted of a mixture of rounded and crushed gravel with 10 nun maximum particle
size. The coarse aggregate water absorption was 0.3 % and bulk specific gravity
2.62, in saturated surface dry condition. Chemical analysis ( XRF method) of the
aggregates was provided by the aggregate supplier. The results of these tests are
shown in Table 3-5.
C4 AF 6.40 - -
Fe - - 0.09
AI - - 0.09
Ca - - 0.11
Mg - - 0.09
K - - 0.30
Na - - 0.09
H2 O - - 0.50
.tl
Chapter 3 !ixperimenfal Programme
Chemical Content (% )
composition Fine aggregate Coarse aggregate
Si02 91.72 92.57
Ti02 0.13 0.25
Al203 3.86 2.22
Fe203 0.96 1.28
CaO 0.50 1.53
MgO 0.34 0.98
Na20 N/A 0.00
K20 2.18 0.93
Cr203 0.025 0.03
Mn304 0.005 0.009
P205 0.02 --
BaO 0.04 0.02
V205 0.007 --
Loss of Ignition 0.95 1.98
42
Chapter 3 Experimental Programme -13
The total cementitious content was kept constant in all the specimens at 350 kg/m3
3
and 450 kg/m respectively. The water / binder ratio of all the specimens was also
kept constant at 0.40. The aggregate / binder ratio was 5.2/ 1 for 350 kg/m3 mixes,
3
4.0411 for 450 kg/m mixes and the percentage of fine aggregate in the total was 35
% for the total aggregate. In the mixes studied in this project in addition to cement
replacement, a proportion of the fine aggregate, amounting to 25 %, was also
replaced by fly ash for Mix-350S and Mix-450S. The final mix proportions of high
volume fly ash concrete considered in this study are listed in Table 3-6.
Table 3-6 High volume fly ash concrete mix proportions ( Unit: kg/m3 )
·0
Chapter 3 Experimental Programme 4-1
The fine and coarse aggregates were used in an air dry condition. Water content was
checked and adjusted before mixing. Due to very low water / binder ratio, different
methods of combining the constituents were tried and the following method was
found to obtain a homogenous mix and without bleeding.
1. The coarse aggregates were first placed in the mixer and half the quantity of
water was than added and the mixer was spun for about a minute.
2. The cement and mineral admixture with the fine aggregates were added, then
mixed 30 seconds.
4. Superplasticizer was then added directly to the mixer and mixed for a further
minute.
All specimens were cast in steel moulds which had been lightly oiled. After
casting, all the specimens were covered by polythene sheet and left in the laboratory.
The specimens were demoulded after 24 hours, and then cured in water for 7 days.
Curing was then continued either in water or in a drying environment. Eight batches
were cast for all the concrete mixtures. Each batch contained six prisms, three prisms
for water curing, and three for curing in a drying environment respectively.
Only one geometry of test specimen was used in the Part I of this research
study, namely, 100 x 100 x 500 mm prisms. The advantage of using 100 x 100 x 500
mm prisms test specimens for this part of the study is that with a single test geometry,
the essential properties of the concrete can be established, thus reducing the amount
of time and effort required for casting various sizes of specimens.
Chapter 3 Experimental Programme
45
From an engineering application point of view of the high volume fly ash concrete,
the mixture proportioning should produce a concrete with adequate flow
characteristics, so that the concrete can be easily placed and compacted. To achieve
this purpose, a relatively large variation in the superplasticizer dosage requirement
was needed, the quantities ranging from 0.9 to 2.5 percent of the total cementitious
content for the different concrete mixtures. The difference in the dosage of the
superplasticizer was mainly due to the different cementitious contents of the high
volume fly ash concrete mixes.
The workability in these studies was determined by the slump test. All the
mixes appeared to have good cohesiveness and no bleeding was detected. With
structural concrete, the aim is to produce a specified workability and 28 day normally
cured strength. The mixes in this study were designed to have a slump of 150 to 180
mm, and the results are shown in Table 3-7.
The normal portland cement concrete mixes had the highest demand of
HRWR of 2.2 % to 2.5 % of the cement content. Mix-350F had the most workable
concrete with the lowest dosage of the superplasticizer. The dosages required for
~5
Chapter 3 Experimental Programme 46
concrete made with Mix-45 OF, Mix-350SF, and Mix-450SF were also quite low.
Mix-350S and Mix-450S had the highest water demand of all concretes with mineral
admixtures, and consequently, higher superplasticizer dosages of 1.6 % - 1.8 % of
total cementitious material were required. The high water demand of Mix-350S and
Mix-450S is most likely due to the higher fly ash content, and the presence of a high
proportion of fine material.
The eight batches of specimens were used for determining the ultrasonic pulse
velocity, dynamic modulus of elasticity, drying shrinkage and expansion at various
ages up to 18 months. At each specific age, the batches were also used for the
determination of air permeability, water absorption, carbonation and compressive
strength following the flexural strength test. After the flexural strength test, each
broken prism was used for a modified cube compressive strength test. The
carbonation test followed the compressive strength test. For each test, three prisms
were used. Table 3-8 shows the details of the tests and the test specimen size used.
The tests are described in detail below.
One of the objectives of this study was to establish whether the incorporation of silica
fume in a high volume fly ash content was beneficial to the early strength
development. Accordingly, flexural and compressive strength tests were carried out
for all of the concretes at 1, 3, 7, 28, 90, 180, 360 days and 18 months after moist or
air curing. At each testing age, three specimens were tested ~ they were removed
from water one hour before testing. For flexural strength, the test procedure was
according to BS 1881 : Part 118 :1983 "Method of determination of flexural
strength" .
~6
Chapter 3 Experimental Programme -17
The dynamic modulus of elasticity of concrete is related to the structural stiffness and
deformation process of concrete structures. The test was determined by the
"Resonant Frequency Tester", made by eNS Electronics Ltd. The instrument
consisted of an "Erudite" electromagnetic exciter unit, an electromagnetic pick-up
unit, a digital counter unit, a variable frequency oscillator and a cathode ray
oscilloscope. The test was conducted by placing the electromagnetic exciter unit and
the pick-up unit, along the longitudinal axis, at the centre of the specimen, against the
100 x 100 mm sides with a path length of 500 mm for each prism. Tests were
performed on the same specimens at intervals of 7 days. The test principle, procedure
and calculation of the value were all according to BS 1881 : Part 209 : 1990,
"Recommendations for the measurement of dynamic modulus of elasticity" .
The purpose of the ultrasonic pulse velocity tests was to look into the effects of the
two different curing conditions with time, and to monitor the development of the
internal structure and microcracking of the high volume fly ash concrete .
.t7
Chapter 3 Experimental Programme 48
The ultrasonic pulse velocity tester used in the present study was the
PUNDIT, manufactured by C.N.S. Electronics Ltd. 54 kHz transducers of 50 mm
diameter were used. The measurements were taken at the two 100 x 100 mm sides,
along the longitudinal axis of the prisms. The measurement intervals at which the
ultrasonic pulse velocity was recorded were the same as for the dynamic modulus of
elasticity test. The test method was in accordance with BS 1881 : Part 203 : 1986,
"Recommendation for measurement of velocity of ultrasonic pulse in concrete".
.t8
Chapter 3 Experimental Programme 49
The volume of a concrete varies with its water content. Drying will cause a volume
decrease, i.e., drying shrinkage, and wetting will cause a volume increase, i.e.,
swelling. Consequently, the rate of removal of water from concrete controls the rate
of shrinkage of the concrete. Reasonably, all factors which affect drying, for example
: temperature, humidity, pressure, air movement, etc., will actually control the
shrinkage.
To study the swelling and drying shrinkage characteristics of high volume fly
ash concrete, water curing and air curing were chosen. After demoulding, all the
specimens had two locating discs bonded to the concrete over a 200 mm gauge
length in the longitudinal direction in the middle of each of the four sides. Initial
readings were taken within an hour of fixing the discs. The measurements were taken
over a period of 18 months at regular intervals of 7 days. For each curing condition,
three specimens were tested, the results are thus the average of twelve readings. The
demec gauge can be read accurately to 8 x 10-6 strain. The swelling and drying
shrinkage were measured according to BS 6073 : Part 1 : 1981, "Specification for
precast concrete masonry units ", Appendix D, "Determination of drying shrinkage".
3-7-5. Carbonation
Carbon dioxide which penetrates the surface of concrete can react with alkaline
components in the concrete, mainly Ca(OH)2. This process leads to a reduction of
the pH value of the pore solution. The depth of carbonation is measured using
phenolphthalein solution [64].
At the specific age of the flexural strength test, the broken surface was
immediately cleaned of dust and loose particles after breaking. Carbonation tests
were carried out on the freshly broken surface by spraying with phenolphthalein
solution. The depth of the carbonation layer from the external surface of concrete to
49
Chapter 3 Experimental Programme 50
the start of the coloration of the concrete was measured as depth of carbonation.
Eight measurements were made on each of the faces of each broken pris~ therefore,
the depth of carbonation reported is the average result of 24 readings.
50
Chapter 4 Test Results and Discussion of Part I 51
CHAPTER FOUR
4-1. INTRODUCTION
A series of concrete mixtures involving two different curing conditions, with and
without fly ash and / or silica fume, with total cementitious contents of 3 50 kg/m3 and
3
450 kg/m , were designed. All the HVF A concrete mixes were made with a water /
binder ratio of 0.4. Because of the very low water content used, large dosages of a
superplasticizer had to be used to obtain the required workability.
The strength of a concrete is its most important property because, after all, concrete
is a construction material. The hardening of the concrete is the result of the hydration
reaction, the subsequent development of bonds in the hydration products, and gradual
increase of the internal strength. Nevertheless, strength usually gives an overall
picture of the quality of concrete, because strength is directly related to the structure
of the hydrated cement paste. Moreover, the strength of concrete is almost invariably
a vital element of structural design, and is specified for compliance purposes.
51
Chapter 4 Test Results and Discussion o[Part J 52
This section presents the results of the strength properties of high volume fly
ash concrete in its hardened form. For purpose of comparison, the strength properties
of normal portland cement concrete are also discussed .
The flexural and compressive strength tests were carried out for all of the
concretes at 1, 3, 7, 28, 90, 360 days and 18 months of moist curing and air curing.
Three specimens were tested at each age, and the results are thus the average of three
measurements. The compressive strength of the concrete was found by the modified
cube test.
The data on the compressive strength development under both curing conditions are
presented in Tables 4-1 and 4-3 and illustrated in Figs 4-1 and 4-3.
The compressive strength development of high volume fly ash concrete under
water curing is shown in Table 4-1 and compared in Fig 4-1.
It can be seen from the table that, the early age compressive strength of
concrete incorporating both fly ash and silica fume, i.e. of, Mix-350SF and Mix-
450SF, is higher in comparison with those of concrete with cement replaced partially
by fly ash alone i.e., Mix-350F, Mix-450F, but lower than the strength of mixes
52
Chapter 4 Test Results and Discussion of Part I 53
containing fly ash as partial replacement of both sand and cement. The increased gain
in strength at early ages of concrete incorporating fly ash and silica fume may suggest
an early start of the pozzolanic reaction due to the use of silica fume, although this
may also be partially accounted for by the filler effect of silica fume resulting in a
denser matrix structure [65].
The data in Table 4-1 show that the compressive strength generally increased
with age for all the HVFA concretes, producing high compressive strength in the
range of 60 to 68 MPa at the age of 18 months. The data also show that the presence
of silica fume had very little effect on the long-term strength of the HVF A concrete.
There are two possible reasons for the higher strength of Mix-350SF and
Mix-450SF concretes at early ages. First, because of its small particle size, silica fume
can act as a filler in the spaces between cement grains. This results in a reduction in
the size of the individual pores and voids in the concrete, although the total porosity
is not affected. Since pores are discontinuities in the cement concrete matrix, reduced
pore sizes require a higher stress to initiate a crack ; thus, the strength is increased
[54]. A second reason for the early high strength of the 350SF, 450SF concrete is
due to the pozzolanic nature of silica fume, which also explains why the 350SF,
450SF concretes become progressively stronger with time relative to the concretes
containing fly ash only.
Mix-350S and Mix-450S containing fly ash as both cement and sand
replacement showed slightly higher strength at all ages than the mixes without and
with silica fume. As mentioned before, early fly ash concrete mixes were based on
low percentage of cement replacements ; these methods are suitable for mass
concrete where early strength development is not a criterion, and they invariably lead
to a lower strength of fly ash concrete [6]. When fly ash is also used to replace part
of the sand in a mix without changing the water and cement contents, the total
cementitious material actually increases and the water / binder ( w / b ) ratio
decreases accordingly. Theoretically, the strength of concrete increases with
decreasing w / b ratio. Therefore, proper use of fly ash to partially replace sand will
be advantageous in strength development. Since the fly ash particle is smaller than the
53
Chapter 4 Test Results and Discussion of Part I 54
sand, the more the sand is replaced, the more the specific surface area is increased.
However, if an excessive portion of sand is replaced by fly ash, the w / b ratio is also
excessively reduced, and thus may, in turn, create workability problems.
When fly ash is used to replace part of the cement and sand simultaneously,
the advantage gained from reducing the w / b ratio may be able to compensate the
loss in early strength due to the presence of fly ash. In other words, using fly ash to
replace both cement and sand partially appears to have a synergistic effect combining
the benefits of both separate replacements.
The results in Table 4-1 also show that the difference in compressive strength
3
between 350 kg/m mixes and 450 kg/m3 mixes is small. Both mixes give 12 to 20
MPa at 3 days, 20 to 30 MPa at 7 days, and 30 to 40 MPa at 28 days. At one year,
the compressive strength ranged from 55 to 60 MPa, and in excess of 60 MPa at 18
months.
Table 4-2 and Fig 4-2 show the rate of gain of compressive strength as a
percentage of the 28 day strength. The results confirm the role of pozzolanic
reaction. At 7 days, compressive strength values ranged from 60 % to 75 % for the
350 kg/m3 concretes, and from 76 % to 81 % for the 450 kg/m3 concretes of their 28
day strength. This was followed by a larger increment after long term water curing.
This increment at 18 months ranged from 157 % to 187 % for the 350 kg/m3 mixes
and from 163 % to 201 % for the 450 kg/m3 concretes compared to the 28 day
strength. These data show that, as expected, the strength development, in general,
increases with age for all the concrete mixtures. The data show that for moist cured
specimens, the higher the compressive strength, the greater was the increase in
strength up to 28 days. The percentage gain in strength of the higher strength
concrete was greater than that of the low strength concrete. However, this is
reversed after about six months ; the stronger the concrete the lower is its subsequent
proportionate gain in moist cured concrete. These results show that when curing
conditions are favourable, the pozzolanic reaction of fly ash continues over a long
period, with the contribution of pozzolanic reaction being especially more prominent
for flexural and tensile strength than for compressive strength [19].
54
Chapter 4 Test Results and Discussion ofPart 1 55
80
o:!
350F 350SF 350S
~ ~ --v- ---B- ~-'7~
,g 60 y-
~ -------~'
~
r.I.l 40
~
'Vi
'"
<I.)
6.20
~
U
'0 180 360 540
1 7 28 90
Age (Days)
80
o:!
450F 450SF 450S
~ -+-- ------ ---*-
·60
~
5
.......
r.I.l
<I.)
40
;>
"Vi
~
6.20
E
0 ~-.;~~;;:-~~~---
U
'0 28 90 180 360 540
1 7
Age ( Days)
55
Chapter 4 Test Results and Discussion of Part 1 56
3 37 40 47 47 42 48
7 60 75 75 78 76 81
200r-------------~========~------------~
350C 350F 350SF 350S
160 .....-- ~ ~ -e--
.,._...-4----
120 ._------_...__._-_........ ...
~
7 28 90 180 360
Age (Days)
200~----------~==========~------------~
450C 450F 450SF 450S
160 - - -+- ----- ------
120
80
40
O'L---~----~~~~------~------~~--~--~~
1 7 28 90 180 360
Age (Days)
56
Chapter 4 Test Results and Discussion of Part I 57
Fig 4-2 also shows that the 28 day strength appears to be some sort of a
threshold stage. Before that mixes with pozzolanic materials appear to give lower
strength than the control concrete, whereas beyond 28 days, the control concrete
gives lower strength than the concrete with the pozzolanic materials.
Under air curing, the compressive strength development of all the concretes
show a similar pattern to that previously observed with water curing, the strength
increasing with increasing age for all the concrete mixtures. However, strength
development is generally lower than that under water curing. All concrete mixtures
had 28 to 40 MPa at 28 days compared to 33 to 41 MPa under water curing. Further,
the data show that for all air cured specimens there is a gradual retrogression of
compressive strength with continued air drying. At one year, the compressive
strength of all air cured concretes was 47 to 62 MPa compared to 55 to 63 MPa for
water curing. All the mixes except Mix-350S and Mix-450S generally showed lower
strength than that water cured specimens. Surprisingly, Mix-350S, Mix-450S showed
least loss of strength due to air drying. Both mixes gave about 57 ~ 58 MPa at 18
months compared to 65 ~ 68 MPa under water curing.
57
Chapter 4 Test Results and Discussion of Part 1 58
80
~ 350F 350SF 350S
::E ~~-B-
,5 60
/)J)
~
r.I.l 40
(\)
;;.
'm
'"
(\)
ts. 20
E
0
C)
'0
1 7 28 90 180 360 540
Age (Days)
80
d
450F 450SF 450S
~ --t-- - - -
,..::' 60
bo
....~
r.I.l 40
(\)
,~
'"'"
(\)
ts. 20
0
E
C)
'0 90 180
1 7 28 360 540
Age (Days)
58
Chapter 4 Test Results and Discussion of Part I 59
200
350C 350F 350SF 350S
160 ._....... --*- -v- -e-
120
?fl
80
40
0 180 360
1 7 28 90
Age (Days)
200
450C 450F 450SF 450S
160 --+- - -
120
~
80
40
0' 28 90 180
1 7 360
Age (Days)
Figure 4-4 Rate of compressive strength as % of 28 days strength-Air curing
59
Chapter 4 Test Results and Discussion of Part 1 60
At one year and 18 months the rate of gain of strength under air curing was
less than that under water curing
3
Table 4-6 Influence of air curing on compressive strength-450 kglm binder
content
(Unit: % )
60
Chapter 4 Test Results and Discussion of Part 1 61
These results emphasize the effect of air curing in the long term. Air curing of
all concrete mixtures exhibits a slight retrogression of strength with continued air
drying. The average compressive strength of air cured concrete at 18 months varied
3
from 76 % to 90 % for the 350 kglm mixes, and from 74 % to 88 % for the 450
kg/m3 mixes of the water cured strength. It is recognized that high volume fly ash
concretes are sensitive to moisture conditions.
The results indicate that the best development of strength gain occurs in moist
curing rather than in air curing. The results also indicate that for a steady increase in
strength for concrete incorporating high volume fly ash content a continuous moist
environment is indispensable. In concretes with low water / cement ratios, once
enough moisture is lost from the concrete, so that the internal relative humidity drops
below about 80 % either by evaporation or self-desiccation, hydration will stop and
strength development will be reduced below its potential, and this reduction will be
greater in the case of high-strength concrete ( low water / cement ratio ) than it is in
the case of low-strength concrete ( high water / cement ratio ) [13]. The loss of
potential strength in the long term is a more widespread and insidious problem, but it
is avoidable. Therefore, it is desirable to provide additional moisture during curing to
ensure maximum hydration.
The ACI Recommended Practice for moist cunng for most structural
concrete, is the time necessary to attain 70 % of the specified compressive or flexural
strength, whichever is less. Various investigators have concluded that the percentage
of strength of dry cured concrete to that of water cured concrete is between 60 % to
90 %, or sometime this varies even more widely.
61
Chapter 4 Test Results and Discussion o[Part I 62
The results from the present investigation of all concrete mixtures under air
curing gave strength values from 70 % to 90 % to that under water curing at 18
months. These data emphasize dramatically the need for early incorporation and a
minimum period of water curing as an integral part of mix proportioning fly ash
concrete.
The data on flexural strength development are presented in Tables 4-7, 4-9, and
illustrated in Figs 4-5 and 4-7. The flexural strength expressed as a percentage of 28
day values are given in Table 4-8 and 4-10 and in Figs 4-6 and 4-8.
The rates of strength development in Tables 4-8 and 4-10 show similar
patterns to compressive strength. Although the higher strength concrete had lower
percentage gain, the flexural strength value in the stronger concrete was higher than
that in the low strength concrete.
The relative rates of development of strength of high volume fly ash concrete
cured in water and in air are illustrated in Tables 4-11 and 4-12. At the end of 18
months, all concrete mixtures exhibited a very minimal reduction in flexural strength
with continued air drying, i.e. about 97 % of the average strength for water cured
concrete. Except Mix-450S which showed a higher loss.
62
Chapter 4 Test Results and Discussion of Part 1 63
~0'6
;;.
'/il
o_~
e'"
S'4
8
12 ----~-- 350F 350SF 350S
CI)
G: -*- ~ -&-
'0
1 7 28 90 180 360 540
Age (Days)
8
<IS
p..
~
0'6
.::
'"'"
CI)
....
e0..4
0
u
"i
.... 2
450F 450SF 450S
~ -+-- - ----*-
G:
'0
1 7 28 90 180 360 540
Age (Days)
63
Chapter 4 Test Results and Discussion o(Part 1 64
200
350C 350F 350SF 350S
160 .-.-..• ~ ~ ----e-
120
--.--.
...~...........-.-.-..
~
80 _....-----------------_.
40
0 90 180 360
1 7 28
Age (Days)
200
450C 450F 450SF 450S
160 --r- - ---*-
120
~
80
40
0' 28 90 180 360
1 7
Age (Days)
Figure 4-6 Rate of flexural strength as % of 28 days strength-Water curing
64
Chapter 4 Test Results and Discussion of Part I 65
8~-------------------------------------------------'
8~-------------------------------------------------r
0:1
p..
:E
Q'6
;;.
"1jl
e'"
~4
U
til
.... 2 450F 450SF 450S
~ -+- -------- ----+-
G:
'0 L_~~~~--.l-...~-===:==;::::::===~~:-----=-:-:-~:-:-::-----:-:-:~
1 7 28 90 180 360 540
Age (Days)
65
Chapter 4 Test Results and Discussion o(Part I 66
3 71 96 84 69 93 82
7 88 97 95 89 96 98
200
350C 350F 350SF 350S
160 _..._.- ~ -4- -B-
120 -...--.-
...
_-_ __
..... ... .--.. .... _--_.----
~
80
40
0 28 90 180 360
1 7
Age (Days)
200
450C 450F 450SF 450S
160
120
~
80
~--
40
0' 28 90 180 360
1 7
Age (Days)
Figure 4-8 Rate of flexural strength as % of 28 days strength-Air curing
66
Chapter 4 Test Results and Discussion o{Part 1 67
Table 4-11 Influence of air curing on flexural strength-350 kglm3 binder content
(Unit: %)
Table 4-12 Influence of air curing on flexural strength-450 kg/m3 binder content
(Unit: % )
These results thus show that the decrease in flexural strength of air cured
concretes is relatively small, and almost negligible. The results indicate that air curing
preceded by moist curing, even for a short period, seems to have beneficial effects on
strength development. These results are in accordance with those reported by Swamy
who studied the effect of initial 7 day water curing on fly ash concrete, and indicated
that early water curing for a minimum period of seven days is an essential ingredient
67
Chapter 4 Test Results and Discussion of Part I 68
for the successful use and full realisation of the strength potential of fly ash concrete
[6].
The relationship between flexural and the compressive strength is generally of the
form: [13]
f =a fc b -------------------------- ( 4-1 )
where f is the flexural strength in MPa, fc is the compressive strength in MPa, and a
and b are constants.
The correlation factor for the two sets of equations is high, and hence, in
practice, a single equation describing the two strength parameters can be used as
given below and shown in Fig 4-11.
68
Chapter 4 Test Results and Discussion o[Part I
10
350F 350SF 350S 450F 450SF 450S
X <> 0 + • I
~
8
~
~
~ .. ~
ts:: 6
. '*
0
x
+~~.
~
VJ
MOIST 0
-a \
3~ 4 •
~x
......
(])
o· o·
~ +
X
+
2
o ~~ __ ~~ __ ~~ __ ~~ __ ~~ __ ~~ __ ~~ __ __
~ L-~
o 10 20 30 40 50 60 70 80
Compressive Strength, MPa
10 ~------------------------------------------~
o
o • o
••
o
x+
DRY
\ o
2
+
/*
o L-~~ __~-L~__~~~--~-L--~~~--~~~
o 10 20 30 40 50 60 70 80
Compressive Strength, MPa
69
Chapter 4 Test Results and Discussion ofPart 1 70
Fig 4-11 is also compared to the data suggested by Swamy & Mahmud and
the relationship obtained from several other authors [9,27,59,60]. Regression analysis
of the data shown in the Fig 4-11 produced the following relationship :
f= 0.31 fe 0.71 (moist + dry) (r = 0.97) ---- ( 4-5) Swamy & Mahmud [59]
10 ~------------------------------------------~
~~~- Swamy
0
0 10 20 30 40 50 60 70 80 90 100
Compressive Strength, MPa
70
Chapter 4 Test Results and Discussion orpar! I 7J
The results in Fig 4-11 show a wide variation in strength development ~ this is
mainly due to the way the HVF A concretes are proportioned, differences in initial
and type of curing, water / binder ratio, and age of testing. It can be observed from
the figure that the author's results do not diverge greatly from a single relationship.
Also, it is very comforting to see, a close similarity to the HVF A concrete data
obtained from Swamy & Mahmud.
The strength characteristics of HVFA concretes can be clearly inferred from Tables
4-1,4-3,4-7, and 4-9. The results also indicate the need for a well-defmed period of
water curing for the strength development of such concrete. It would then be
interesting to compare the strength development of HVF A concretes with that of
normal portland concrete at different ages, and under different curing conditions.
Table 4-13 represents the strength data for both ope concretes under both
types of curing conditions. The rate of ope concrete strength development as a
percentage of 28 day strength development compared to all HVF A concretes is also
shown in Figs 4-2, 4-4, 4-6, and 4-8.
71
Chapter 4 Test Results and Discussion o[Part I 72
concretes at an early age, but after 28 days, it is lower than that for HVF A concretes
under the same conditions.
From the results, it can be seen that under continued moist curing condition,
the percentage increase in compressive strength from 28 days to one year is 17 % ~
72
Chapter 4 Test Results and Discussion ofPart I 73
The results also indicate that under air curing, some loss of strength occurs
with OPC concretes with age, whereas HVF A concretes are more likely to maintain
hydration process with a gradual rate of increase. At the end of one year, a slight
retrogression of strength on HVF A concretes with air curing was similar to that of
normal OPC concrete, but the percentage gain on HVFA concrete was still higher.
For example, the difference between 350SF and 350C, 450SF and 450C at one year
with air curing was 69 % and 37 % on compressive strength, and 34 % and 6 % on
flexural strength, respectively.
These results indicate that even under continuous air drying, the HVF A
concretes are able to continue hydration process, although at a slower rate.
The superiority in the rate of later age strength development of the HVFA
concretes is clearly seen, even though had lower strength than that of OPC concretes.
This increase in the rate of strength development of the HVF A concretes over that of
OPC concretes was about 40 % of the average values on compressive strength, and
30 % on the flexural strength. The incorporation of fly ash, silica fume and a
superplasticizer in HVF A concrete is thus clearly beneficial to its long term strength
development.
73
Chapter 4 Test Results and Discussion of Part 1 7-1
The effect of total binder content on concrete strength development under both types
of curing conditions is shown in Figs 4-12, 4-13, 4-14, and 4-15. The results suggest
several aspects related to the role played by the superplasticizer and the binder
content. For purposes of comparison, the results for the normal portland cement
concrete are also included in the above figures.
The results show that the total cementitious binder content does not have
much effect on the value of strength. In general, concrete with 450 kg/m3 is only
3
marginally better than 350 kg/m mixes under both types of curing conditions.
In the current study, the compressive strength of 450C under both curing
conditions was somewhat lower than that of the concrete made of 350C. Previous
discussion has shown that the concretes of 450C had superplasticizer demand of2.5
%, higher than that for the concrete of 350C. The lower strength of concrete with
increasing dosage of superplasticizer is due, at least in part, to the greater fluidity and
the accompanying segregation and bleeding that occur in these concretes. These
results conform to the explanation of Goldman and Bentur [66] who used
progressively more superplasticizer as the silica fume content of their concrete was
increased. It appears that, although the workability of their concrete was enhanced,
the concretes became too fluid at the higher superplasticizer contents, resulting in a
reduction in strength similar to that observed in this study.
The results stress two points : that for the HVF A concrete to have a certain
specified strength, it is important to choose a suitable cementitious binder content
with the object of producing, as economically as possible, concrete of the required
strength. In addition, the selection of mix proportions has to take into account the
method of transporting the HVFA concrete, especially if pumping is envisaged.
74
Chapter 4 Test Results and Discussion o{Part I
100 100
'" '"
~ 80 ~ 80
350F450F
+
-!c oS
bJ)
:::
*
60 E 60
E
r/.l r/.l
u u
;> ;>
"!iJ 40 "!iJ
e'" '"
8
a- 20
0
1 7
I 35OC450C
28
I
90 540
! 7 28 90 540
Age ( Days) Age (Days)
100 100
~'"
<IS
350SF450SF 350S450S
~ 80 ~ 80 -e- .....
~g ~ :::
60 u 60
r/.l -=
r/.l
u u
;> ;>
"1jJ 40 "Til 40
e'" e'"0.
a-
0
80 20
U U
0
7 28 90 540 1 7 28 90 540
Age (Days) Age (Days)
100 100
o:s <IS
350F 450F
~ 80 ~ 80 +
tc ~c *
g 60 g 60
r/.l r/.l
u <I)
;> ;>
"1jJ 40 "a=J 40
en
e'" e0.
a- 20 I 350C450C I 8
u0 u0
0 7 28 90 540
1 7 28 90 540
Age ( Days) Age (Days)
100 100
o:s o:s
350S 450S
~ 350SF450SF ~
t:::
80 ~ --- ~
80 -e- .....
:::
u 60 <I)
60
-~
-=
r/.l -=
r/.l
<I) <I)
;>
;>
"1jJ 40 "1jJ 40
en en
e /~/ e0.
S' 20 ~/
80 20
U
0
~..v' u
0
7 28 90 540 1 7 28 90 540
Age (Days) Age (Days)
75
Chapter 4 Test Results and Discussion orPart 1
10 10
/\ = 350F450F
~ 8
/ ../
/"" \ . ~
..g
8
* +
..g
gp 6 /;l()
6
g g=
00 00
4 4
e 1~
~
~
(I)
2 I 350C450C I ~ 2
'0 0
1 7 28 90 540 1 7 28 90 540
Age ( Days) Age (Days)
10 10
3508F4508F = 35084508
~ 8 ......
~
~
~ 8 -e-
~g 6 ~g 6
00 00
e 4
e::1
4
~ 2
~
~ 2
~
'0 0
1 7 28 90 540 1 7 28 90 540
Age (Days) Age (Days)
Figure 4-14 Effect of total binder content on flexural strength (Water curing)
10 10
350F 450F
~ 8 =
0... 8 +
~ ~ *
~ 6 ~ ~
6
.....~ .....
00
00
4
e 4
1~
~ 2 I 350C45OC I ~ 2
~
'0 0
1 7 28 90 540 1 7 28 90 540
Age (Days) Age (Days)
10 10
35084508
350SF4508F
~ ....... =8
0... -e- ......
~
~~
~ 6
J''<:' • • .....~
00
4
:<V
-"'" ,-~
1~ 2
'0 0
1 7 28 90 540 1 7 28 90 540
Age (Days) Age ( Days)
Figure 4-15 Effect of total binder content on flexural strength (Air Curing)
76
Chapter 4 Test Results and Discussion of Part 1 77
The results of the present study emphasize the importance of the characteristics of
concrete incorporating high volume fly ash. Compressive strength is in the range 30
to 40 MPa at 28 days, and 40 to 60 MPa was achieved for all the HVF A concretes at
the age of 90 days. In published research, a higher amount of cementitious binder
content, and lower water / binder ratios have been used to produce HVF A concrete
systems. Swamy and Mahmud [59], for example, produced HVF A concrete with
cube compressive strength of23.7 to 63.3 MPa and prism flexural strength of2.6 to
4.5 MPa at 28 days, using a total cementitious binder content of 266 ~ 466 kg/m3 and
water / binder of 0.32 ~ 0.62. Malhotra and Giaccio used [61] cementitious content
of 375 kg/m3 with water / binder 0.31, and the cylinder compressive strength of
HVFA concrete was 32.5 MPa at 28 days, and 41 MPa at 91 days. Langley et al [60]
developed HVF A concrete cylinder compressive strength of 23 to 57 MPa, and 38 to
75 MPa, and 6 to 6.9 MPa, and 8.9 to 9.6 MPa flexural strength at 28 days and 90
days respectively. The HVF A concrete was produced using a total binder content of
225 ~ 400 kg/m3 with water / binder 0.28 ~ 0.49. In Electric Power Research
3
Institute ( EPRI ) research [9], a cementitious content of 365 kg/m with a water /
binder 0.33 was used, and the cylinder compressive strength of HVF A concrete
varied from 27.8 to 42.2 MPa at 28 days, with 50 MPa at 91 days. The prism flexural
strength reached 5.2 ~ 6.5 MPa at 91 days. Table 2-5 in Chapter 2 also shows a
comparison of strength development obtained from other investigators.
Cement paste or concrete cured continuously in water from the time of placing
exhibits a net increase in volume and an increase in mass. This swelling is due to the
absorption of water by the cement gel ; the water molecules act against the cohesive
forces and tend to force the gel particles further apart, with a resultant swelling
77
Chapter 4 Test Results and Discussion of Part I 78
pressure. In addition, the ingress of water decreases the surface tension of the gel,
and a further small swelling takes place.
presented, and the results discussed in terms of the different replacement materials
and the total cementitious content. Under water curing condition, the swelling each
specimen was measured on all four sides, and the results are thus the average of
twelve readings.
Typical swelling test data for all HVF A concretes are shown in Table 4-14
and illustrated in Figs 4-16 and 4-17.
The results show that the final swelling varied over a narrow range, between
100 and 125 micro-strain depending on the total cementitious content. The swelling
initially increased rapidly, after which there was a steady increase with continued
water curing. Between 9 months and one year, there was some sharp increase in
swelling of all the HVF A concretes, beyond which the rate of swelling became more
stabilized till the end of test period. In general, the concrete mix with 450 kg/m3,
showed lower swelling. At the end of 504 days, the swelling value varied from 105 x
10-6 to 125 X 10-6 for the 350 kg/m3 mixes and 100 x 10-6 to 120 X 10-6 for the 450
kg/m3 concretes.
78
Chapter 4 Test Results and Discussion of Part I 79
The results also show that the HVFA concretes incorporating small amount
of silica fume benefited marginally to suppress swelling. This finding implies that
silica fume acted on the one hand as an intensifier of pozzolanic reaction, and on
other, as an activator of fly ash, while the combined influence of silica fume and fly
ash controlled swelling to an even greater extent.
In view of the swelling of this type at an early age, it can be related to the
hydration process in concrete. According to the swelling theory [67] swelling occurs
because of wetting of the ettringite particles that are of colloidal size ( < 1 ~m in
length ), from a topochemical growth of ettringite pushing against solid material in
the concrete. The early hydration reaction is located at the acceleration stage of the
calorimetric curve which is due to the formation of ettringite as the hydration of C3A
and sulphate ion occurs, and the extra pressure of the inner part in C3 S grains
accelerates the concentration of calcium, in which a large amount of CH crystal as
well as C-S-H gel will form.
79
Chapter 4 Test Results and Discussion of Part I 80
1~ r----------------------------------------------,
120
-
\0
I
90
.~
.j:j
00
.---~
Q
( l)
60
~
00
30
350F 3508F 3508
)( 0 0
o~~~~~--~~~~~--~~~~~~--~~~~~
150 .----------------------------------------------,
120
-
\0
I
Q 90
·ca
.j:j
00
.---~
Q
( l)
60
~
00
30
450F 4508F 4508
•
o~~-L~~L-~~-L~~--L-~~-L~--~~~~~~
80
Chapter 4 Test Results and Discussion of Part I 81
150
350F 450F
x
120
\0
I
0
....-4
0 90
'ed
~
00
.-..........
OJ:)
0
Q) 60
~
00
30
o~~-L~~~ __ L-~~~~~--L-L-~~~~~--~
81
Chapter 4 Test Results and Discussion of Part J 82
150
350SF 450SF
e
120
\0
I
0
,.......;
~
90
·a /~
~
r/J
bI)
~
......
.~
...... 60
(I)
~
r/J
30
o~~~~~--~~~~~--~~~~~~--~~~~~
150
350S 450S
e •
120
\0
I
0
,.......;
90
.S
t\l
~
r/J
bI)
~
......
.~
..- 60
(I)
~
r/J
30
o~~-L~~ __ ~~~-L~--L-~J-~~~--L-~-L~~
82
Chapter 4 Test Results and Discussion ofPart I 83
Drying shrinkage of concrete is caused by the contraction of the calcium silicate gel
in the hardened paste as the moisture content is reduced. As concrete dries, the water
lost first comes from the capillaries and causes little shrinkage. After the capillary
water has been lost, the water is removed from the cement gel pores and noticeable
drying shrinkage occurs. While many factors determine the magnitude of drying
shrinkage, the materials that mostly affect drying shrinkage are water, the
cementitious paste, admixtures and aggregates [68].
In this section, the results of the tests on drying shrinkage of HVF A concretes
are presented. The drying shrinkage specimens were initially moist cured for 7 days,
after which they were transferred to the ambient environment in the laboratory and
monitored over a period of 504 days at regular intervals of 7 days. The effect of
variations in different replacement methods and total cementitious content was also
investigated. The experimental drying shrinkage data are evaluated with the aid of
Ross' formula [13]. The relationship between the drying shrinkage and the swelling is
also reported.
Drying shrinkage values are expressed as linear strain as a function of age and shown
in Table 4-15 and illustrated in Figs 4-21 and 4-22.
At early ages, the HVF A concretes have not fully matured, and the high early
drying shrinkage strain obtained at this stage were expected. A high proportion of the
drying shrinkage strain for all the air dried HVF A concretes occurred during the
initial 28 days, after which the shrinkage gradually increased with increase in age and
then became somewhat stabilized with time until the end of the test period.
83
Chapter 4 Test Results and Discussion ofPart I 84
Compared with all HVF A concretes, it can be seen that the drying shrinkage
strain results obtained in this study are relatively high. At the end of the test period,
the drying shrinkage strain of all the HFVA mixtures varied between 550 and 640
microstrains. According to the data obtained in this investigation, mixtures with
cement replaced partially by fly ash alone, i. e. mixes, 350F, 450F showed greater
resistance to drying shrinkage. Mixes 350S and 450S with large amounts of fly ash in
the concretes showed slightly higher drying shrinkage strain, whereas mixes 350SF
and 450SF showed the highest drying shrinkage strain values at the age of 504 days.
The addition of silica fume did not thus seem to restrain significantly the drying
shrinkage of air dried HVFA concretes. In contrast, the swelling of HVF A concretes
stored in water indicated that concretes incorporating of small amount silica fume,
350SF and 450SF, showed a significant reduction in swelling, and had good ability to
suppress swelling.
The reduction in swelling could also be attributed to the lower water / binder
ratio of the HFVA concrete, and the changed pore size distribution of the matrix,
thereby slowing the rate of swelling. Normally, swelling and drying shrinkage are
related - the higher the swelling reduction, the greater its the drying shrinkage [69].
200 r-------------------------------------------~
O~--------------------------------~
'D
I
o
-" (100)
~
j (200)
00 (300)
I:lI)
.S
g (400)
(500)
(600)
200 r - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - ,
o
-" (100)
Q)
~ (200)
]
00 (300)
I:lI)
.S
g (400)
(500)
(600)
85
Chapter 4 Test Results and Discussion ofPart 1 86
Figs 4-23 and 4-24 represent the swelling / shrinkage characteristics for
HVF A concrete from the present study. It can be seen from the figures, that the
higher the swelling of 350F and 450F, the lower is its drying shrinkage strain. To
investigate the relationship between drying shrinkage and swelling of HVF A
concretes, a plot of drying shrinkage data versus swelling at approximately the age of
504 days is presented in Fig 4-25. Regression analysis of the data shown in the figure
produced the following relationship :
The figure indicates that the good inverse correlation between drying
shrinkage and swelling is linear.
Table 4-16 shows typical data from the present study of M. I. P. test on all
HVF A concrete air dried specimens. It can be seen from the table, that at the age of
18 months, the percentage gain in coarse pore size of concretes with cement replaced
partially by fly ash alone, 350F and 450F, is higher in comparison with those
incorporating both fly ash and silica fume concretes. The drying shrinkage behaviour
of HVF A concretes in the present study is explained in terms of coarse and fine pore
size with increase in coarse pore size resulting in a decrease in drying shrinkage and
increase in fine pores producing increased drying shrinkage. The results on drying
shrinkage thus confirm those observed in published literature [71].
86
Chapter 4 Test Results and Discussion of Part 1 87
200
100 ®
'9 ~§ <>
0 0
.-4
d'
.~ (100)
350F
en
)(
~ (200)
350SF
o
] (300)
-.S
en
.-.
.-.
(400)
X
350S
o
Q) 0
~ (500) <>
(600)
(700)
0 112 224 336 448 560
Age (Days)
200 ~--------------------------------------------~
t
\0
I
100
:.
o o ~--------------------------------------------~
.-4
.~~ (100) +
450F
en
Q)
~ (200)
450SF
~
-.S
] (300)
en 450S
•
(400)
.-.
.-.
Q)
~ (500)
+
(600) •
(700) L....._--'-_---1.._ _. L . . . . - _ - L - _ - - L_ _L - - _ - - ' - _ - - - L . ._ _L..==-...J
87
Chapter 4 Test Results and Discussion of Part I 88
6W r-----------------------------------______~
640
-
1.0
I
o
o
o
00
y = 879.686-2.5860 x
r= - 0.9799
560
Table 4-16 Typical % of meso pore of high volume fly ash concrete under air
curing
(Unit: % )
88
Chapter 4 Test Results and Discussion of Part 1 89
Although the HVF A concrete exhibited high drying shrinkage strain values
early in its life, these drying shrinkage strain levels are either comparable to or in a
manner similar to those observed by Malhotra and his co-worker [9,43]. They
obtained a range of drying shrinkage from 400 ~ 600 microstrains at the end of 448
days with similar concretes under essentialy the same test conditions.
The effect of total binder content of all HVF A concretes on drying shrinkage
IS compared in Figs 4-26, 4-27, and 4-28. The figures show that the total
cementitious content has only a marginal influence on the values of drying shrinkage.
As expected, the 450 kg/m3 mixes had a slightly higher drying shrinkage strain than
with 350 kg/m3 . The higher drying shrinkage of 450 kg/m3 mixes is probably due to a
combination of two factors - a higher binder content and superplasticizer dosage
[42]. In addition, at a constant water I binder ratio, drying shrinkage increases with
an increase in the cementitious content because this results in a large volume of
hydrated cementitious paste which is liable to drying shrinkage [69].
200 ~------------------------------------------~
350F 450F
100 )(
O~-------------------------------------------i
\0
I
o
.......~ (100)
Q)
bO
~ (200)
]
00 (300)
00
a
.S
(400)
(500)
(600)
(700) L-..l.----L--L--L.--l...----1--''---.L-..-J.--L--L----'-----'----'---L...-.1.--.-'----'---'----'
89
Chapter 4 Test Results and Discussion of Part 1 90
200 r-------------------------------------------~
350SF 450SF
100 o
O~-------------------------------------------~
o
......" (100)
(!)
E~
(200)
]
r:Fl (300)
gf
a
......
(400)
(500)
(600)
200 r--------------------------------------~
350S 450S
100 o •
o~------------------------------------------~
o
......" (100)
(!)
bO
~ (200)
~
.\::
..e
r:Fl (300)
gf
a
......
(400)
(500)
(600)
90
Chapter 4 Test Results and Discussion ofPart I 91
Many kinds of equations including ACI, CEB formulae, Meyer, and Bazant have
been developed to predict the drying shrinkage of conventional concrete. In the
analysis of the drying shrinkage of the HVF A concrete under investigation, a
hyperbolic expression suggested by Ross has been used [69].
t
c = -------------------- ( 4-10 )
a+bt
where c is the drying shrinkage of HVFA concrete in microstrain, t is the time after
initial readings for drying shrinkage in days, and a and b are constants.
Figs 4-29 and 4-30 show the curves for determining the constants "a" and "b"
for all HVF A concretes undergoing drying shrinkage. Table 4-17 shows the constants
derived from these results. In the same table, typical results of the measured and
predicted values of drying shrinkage at various ages for all HVF A concretes are also
shown.
From the results, during the first 3 months, the drying shrinkage strains
obtained through Ross's formula are substantially higher compared to experimental
results. However, as can be seen from the table, there is good agreement between the
measured and predicted values, especially for ages greater than 6 months. The drying
shrinkage strains measured on all HVF A concretes are comparable to predicted
values, i. e., about 95 % of the average percentage of the latter.
91
Chapter 4 Test Results and Discussion ofPart 1 92
0.8
(])
~
~ 0.6
]
00 /
--
~ 0.4
<
o ~~~~~--~~~~--~~~~--~~~~~--~~~
--
~ 0.4
<
0.2
o L-~-L~ __L-~-L~~L-~~-L~--~~~~--~~~~
o 56 112 168 224 280 336 392 448 504 560
Age (Days)
92
Chapter 4 Test Results and Discussion of Part I 93
Table 4-18 shows a comparison of drying shrinkage from the present investigation
with the results of other researchers [9,29,44,59,72]. Data from the table show that
93
Chapter 4 Test Results and Discussion of Part J 94
there is a wide range of data on drying shrinkage reported on fly ash concrete mainly
due to the different concretes tested, differences in type of curing, and age of testing.
When all the fly ash concretes under various curing regimes are considered, the
drying shrinkage varies from 385 to 642 microstrain. From these data, it is obvious
that it is impossible to make in any general conclusions covering all types of fly ash
concretes. It is worth to point out, however, that the results of drying shrinkage
obtained in the present study are similar to those reported by EPRI and Malhotra ;
they are, however, much higher than the results reported by Swamy. The reason for
this difference is probably due to the mix proportions and the water / binder ratio.
Researchers Fly ash W/C Cube St. Curing Age at last Drying
content (28d) reading shrinkage
(% ) MPa conditions (days) [( microstrain )
0046
Cripwell 35 44.5 20°C + 70 % R. H 300 485
[29] 0.35
" 53.0 " " 400
0043 26.8* 23°C + 70 % R. H 91 500
Ward 40
0041 26.0*
[72] 50 " " 580
0.32 35.0* 7 wet + air curing 448 500
Malhotra [44] 58
0.32~0.62
Swamy 50 24~63 Outside 550 385 ~ 422
[59] "
II
"
II
"
II
94
Chapter 4 Test Results and Discussion of Part I 95
There is now absolutely no doubt that the use of high strength and high
performance concrete in construction will continue to grow. The elastic deformation
of high strength / high performance concrete is of particular importance. Because the
modulus of elasticity of the strong hardened cement paste and of the aggregate differ
less from one another than in medium strength concrete, the behaviour of high
strength / high performance concrete is more monolithic, and the strength of the
aggregate-matrix interface is higher.
In this section the main results deal with the characteristics of dynamic
modulus of elasticity in different mixes and different binder contents concrete under
both types of curing conditions. For purpose of comparison the results of the normal
ope concretes are also discussed. A comparison of the correlation between the
dynamic modulus of elasticity and compressive strength of HVF A concrete is also
reported.
95
Chapter 4 Test Results and Discussion o(Part 1 96
Under both curing conditions at various ages, the dynamic modulus of elasticity of
HVF A concretes with both cementitious contents is illustrated in Figs 4-31, 4-32, 4-
33, and 4-34. Tables 4-19 and 4-20 summarize typical data of all HVFA concretes
under water and air curing, respectively.
96
Chapter 4 Test Results and Discussion o[Part I 97
The results indicate that the modulus increased rapidly during the first few
days, it then increased slightly with further curing, and finally reached a stable value
with continuing hydration reaction. Data from the tables show that the modulus of
elasticity of moist cured specimens is consistently higher than that of the dry cured
specimens. The results also show that the total binder content is a predominant factor
in determining, the value of dynamic modulus. Concrete with 450 kglm3 was superior
3
to those with 350 kglm under both types of curing. Irrespective of the cement
replacement methods, the difference between 350 kglm3 and 450 kglm3 mixes under
both curing conditions is seen to be identical. At the age of 504 days, the difference
in modulus of elasticity values between 350 kglm3 and 450 kglm3 mixes is about 20 -
25 % for both curing conditions. The total binder content generally controls the
concrete making properties of the aggregate, because it determines the amount of
binder paste needed to cover the aggregate surface. A high binder content of
concrete means that it can be significantly enhanced via the physical mechanisms of
improved packing of hydration products, so that the strength of the matrix is high.
Because hydration starts at the surface of the binder particles, it is the total surface
area of binder that represents the material available for hydration. Therefore, the
strength between the coarse aggregate and the matrix, and the elastic properties of
aggregate have considerable influence on the dynamic modulus of elasticity of
concrete.
The densifYing effect of fly ash and silica fume are demonstrated in the data
presented here. However, Mix-450SF concrete with silica fume had a higher dynamic
modulus of elasticity, whereas Mix-450S with a large amount of fly ash had a slightly
lower modulus than the concrete made of 450F.
97
Chapter 4 Test Results and Discussiun of Part I 98
70 ,--------------------------------------------
20
350F 350SF 350S 450F 450SF 450S
)( o o •
10 ~~~~~~~ __ ~~~_L~~~~L_~~_L_L~~
70 ~------------------------------------------~
60
20 450S
350F 350SF 350S 450F 450SF
)( o o •
10 L-~L_~~~~~_L~_L~~~~~~~~~~
98
Chapter 4 Test Results and Discussion ofPart I 99
70 70
---ca 60 ca
6 V- @) 60
'-" '-"
fJl 50 l------.-.-.--.-..----.-- fJl 50
.§ .§
"d 40 "d 40
~ ~
.~ 30 -~ 30
~20 [ 3:~45OC [ S20
Cl
10 10
0 112 224 336 448 560 0 112 224 336 448 560
Days Days
,-... 70 ,.-... 70
<:d ca
6 60 6 60
'-" '-"
~
;3
~ 40
50
- 4050
~
;::1
"d
Alli1~~f3CkJC 8-60 0 0 0 0 0 0 0 O-£)
r o
20
10
0 112 224
Days
336 448 560
r o
20
10
0 112 224
Days
336 448 560
Figure 4-33 Effect of total binder content on dynamic modulus ( Water curing)
,-...
70 ,-...
70
<:d <:d
6 60 @) 60
'-" '-"
50 .-..__ --..__..... _....-- ...__.....__ ....-.
fJl fJl 50
.§ .§
"g 40 "d 40
;:;s ~
r o
20
10
0 112
[ 350C 450C [
224
Days
336 448 560
-~ 30
S20
10
0 112 224
Days
336 448 560
~
70 ,.-...
70
<:d ca
6 60 6 60
'-'
~ 50
"d 40
-
f ........................ • • • •
'-'
.a 40
fJl
't:I
50 ,... - --""
l~
~ I~",
~
.L'. ~~~
r
~
A/,-/\..6.+'>
!? ~
o
-I 30
5
20 !> I 35!F45!F I 20 13~S4!S I
10 10
0 112 224 336 448 560 0 112 224 336 448 560
Days Days
Figure 4-34 Effect of total binder content on dynamic modulus ( Air curing)
99
Chapter 4 Test Results and Discussion ofPart I 100
From the Table 4-19, the dynamic modulus of elasticity of HVF A concretes
ranged from 36 to 51 GPa at 28 days, from 39 to 55 GPa at 91 days, and from 45 to
60 GPa at 18 months for both cementitious contents. Considering the levels of
compressive strength of HVF A concrete at these three ages, the dynamic modulus of
elasticity values are rather high. As mentioned in the previous section, the generally
high dynamic modulus of elasticity values are due to the unhydrated fly ash particles
acting as fine aggregate. As for the chemical reaction of silica fume, because of high
surface area and high content of amorphous silica in silica fume, this highly active
pozzolan reacts more quickly than ordinary pozzolans [73].
From the Fig 4-32, air dried specimens showed slight retrogression of
dynamic modulus of elasticity and show similar trends to changes in strength when
concretes are exposed to a drying environment. As in the strength development
studies, the effect of curing conditions on dynamic modulus of elasticity is significant.
At the end of 504 days, the dynamic modulus of elasticity of HVF A concretes under
air drying ranged from 34 to 46 GPa compared to 45 to 60 GPa for water cured
specimens. Concretes made with both fly ash and silica fume exhibited slightly higher
values than concrete made of350F and 450F.
364 73 76 81 72 77 77
504 74 76 80 73 76 78
Remarks : Water cured specimen is 100 %
100
Chapter 4 Test Results and Discussion o/Part 1 101
101
Chapter 4 Test Results and Discussion o[PartJ 102
189 95 95 94 93 95 94
364 95 95 95 93 95 95
504 96 96 96 96 96 96
From the results, it can be seen that, under continuous moist cunng
conditions, the percentage increase of dynamic modulus of elasticity of normal OPC
concrete from 28 day to one year is 5 % to 6 %. The corresponding average increase
for the HVFA concrete was 13 % to 28 %. The percentage increment of dynamic
modulus of elasticity of HVF A is thus about 3 - 4 times that of the normal OPC
concrete. As water curing continued, the HVF A concrete appeared to undergo some
102
Chapter 4 Test Results and Discussion ofPart I 103
more hydration reaction than the ope concrete although equivalent mechanical
properties were never reached.
At the end of 504 days, the average percentage loss on dynamic modulus for
dry cured specimens of HVF A concrete is quite low, about 4 %, after the 28 day
value. The decrease in modulus of HVF A concrete is thus very small and almost
negligible. It is worth recalling the effect of prolonged air drying on strength
development. With a short period of seven days moist curing, it is possible not only
make the full achievement of the strength potential of HVF A concrete, but in effect
also preserve the complete development of its modulus of elasticity [6].
When high volume fly ash / high strength concrete began to be developed,
efforts were made to see whether the existing relationships could be used to predict
high volume fly ash concrete modulus of elasticity or new formulae had to be
developed.
103
Chapter 4 Test Results and Discussion of Part 1 104
Regression analysis of the data shown in these figures produced the following
relationships :
3
ED = 14.07 fc°.3O (350 kg/m mixes) (r = 0.98) --------- ( 4-12 )
30
ED = 21.50 fco. (450 kg/m3 mixes) (r = 0.96) --------- (4-13 )
Irrespective of the total cementitious content, for the wet cured concrete, the
dynamic modulus of elasticity, ED, increases approximately with the cube root of its
curing compressive strength, fc . It is clear that dynamic modulus of elasticity can be
accurately predicted from compressive strength, for moist cured specimens. Each
individual equation produces a high correlation factor between the two parameters
under water curing and there is a significant correlation at the 99 % confidence level
between dynamic modulus of elasticity and compressive strength of HVFA concrete.
104
Chapter 4 Test Results and Discussion of Part I
70 ,------------------------------------------
3
350 kg / m 450 kg / m3
o •
60
r:o:s
fj
~' 50 • •
'"S
"0
0 ••• o
~
u •• o
.~ 40
~ • o
6' o
o
o 0
30
20 ~~~--~~--~~--~~~--~~--~~--~~~
o 10 20 30 40 50 60 70 80
Compressi ve Strength, MPa
70 .---------------------------------------------~
3 3
350 kg / m 450 kg / m
o •
60
•• •
. ... ••
• •••
30
o 10 20 30 40 50 60 70 80
Compressive Strength, MPa
105
Chapter 4 Test Results and Discussion o{Part I 106
On the basis of the above results, the relationships between dynamic modulus
of elasticity and compressive strength are curvilinear since the dynamic modulus of
elasticity is more dependent on the moisture gradient.
While many results of tests for modulus of elasticity of concrete have been published,
there are only a few where the dynamic modulus of elasticity on HVF A concrete have
been tested and its relationship reported. Most of the researchers [9,27,43,59] have
presented the results of static modulus of elasticity. For this reason, Eqn. 4-16, as
suggested by the British Code for design of concrete structures, CP 110 : 1972 is
also used in the present investigation. The equation may be more convenient to
estimate the static modulus of elasticity, Ec, from the measured dynamic modulus of
elasticity, ED.
Form the figure, good agreement was found between the predicted and
3
measured values for the modulus of elasticity ofHVFA concrete made of350 kg/m
mixes, but, it was lower than the reported by EPRI [9] and Swamy [59]. However,
the levels of compressive strength of the HVF A concrete at 28 and 504 days, and the
modulus of elasticity of 450 kg/m3 mixes are rather high. At the end of 18 months,
the test results are higher than the ACI Building Code 318-89 calculated values by
about 43 ~ 49 % for the 450 kg/m mixes.
3
106
Chapter 4 Test Results and Discussion of Part 1 107
lEe = 9.10 feo. 33 ( CP 110 ) 36.1 35.6 36.6 36.1 36.4 37.0
lEe = 10.95 feO. 3 (Swamy) 37.9 37.5 38.3 37.9 38.1 38.8
lEe = 8.79 fe°.39 (EPRI ) 44.0 43.4 44.7 44.1 44.5 45.4
( *Ee - Ee ) / Be (%)
107
Chapter 4 Test Results and Discussion o[Part I 108
70 r---------------------------------------
Author Author
ACI318-89 CP 110 Swamy EPRI
350 kg/n? 450 kg/of
60 • o o •
Author
•
Swamy
20 CP 110
o 10 20 30 40 50 60 70 80
Compressive Strength, MPa
Figure 4-37 Comparison of predicting modulus elasticity and published data
The ultrasonic pulse velocity through concrete is the outcome of the time taken by
the pulse to travel through the hardened cement paste and through the aggregate.
The pulse velocity change is useful information for understanding the development of
the inner structure of the hardened concrete. The test is , therefore, useful to detect
cracking, voids, deterioration due to frost or fire, and the uniformity of concrete in
similar elements.
108
Chapter 4 Test Results and Discussion o[Part I 109
their values will be described. The experimental ultrasonic pulse velocity data on
HVF A concretes and normal ope concrete are also compared. A relationship
between the ultrasonic pulse velocity and the dynamic modulus of elasticity is also
described.
The variation of ultrasonic pulse velocity of HVFA concretes with both binder
contents under the two curing regimes is illustrated in Figs 4-38 and 4-39. Some
typical data of pulse velocity of HVFA concrete under water and air curing are
shown in Tables 4-27 and 4-28, respectively.
These results show that the ultrasonic pulse velocity curves can be divided
into three stages for all the HVFA concretes. In the first stage, the pulse velocity
increased rapidly for about one week, after which the reading showed an obvious
change under the two curing regimes. The second stage was a transition stage, and
the pulse velocity in this section exhibited a gradual increase or decrease about for 3
months. This changeable stage appears over the period between the end of the early
hydration and the pozzolanic activity period. There was seen during this period from
the preceding sections, a large variation of the swelling due to the formation of
ettringite ; growth of ettringite at early ages becomes more efficient in producing
swelling in the HVFA concretes and therefore, a little unstable stage of pulse velocity
development between the early hydration and the pozzolanic activity period is
understandable [77].
The third stage is a steady and stable phase with a gradual increase in pulse
velocity. The readings remained relatively constant with age at this stage as the
curing continued. From the data it can be inferred that the steady state implies that
the hydration products fill pore spaces and that the hydration reaction continues.
During the pozzolanic activity, the growth of hydration products continues to
progress gradually, and so, the pulse velocity reaches a stable progressive state.
109
Chapter 4 Test Results and Discussion of Part I 110
5 ,---------------------------------------~
4.75
(,)
(])
~ 4.5
]
~
·0 4.25
o
.......
~
(])
-Ep. 4
Age (Days)
Figure 4-38 Variation of pulse velocity on HVFA concrete under water curing
Data from the results indicated that the ultrasonic pulse velocity was the
highest in water cured concretes. From Fig 4-38 it is seen that the presence of silica
fume in Mix-350SF and 450SF helped these concretes to develop a denser internal
structure from an early age. Mixes 350S and 450S with large amounts of fly ash also
110
Chapter 4 Test Results and Discussion of Part I 111
developed a dense structure and their pulse velocity increased very rapidly from early
ages, also all they showed marginally better performance than the silica fume mixes,
but by about six months, all the mixes appeared to have more or less the same value
of ultrasonic pulse velocity. Beyond the age of six months, the pozzolanic reaction
activated by the fly ash and silica fume with the hydration of cement reduced the gap
between the different mixes. The long term pulse velocity developed for all the
HVFA concretes ranged from 4.74 Ian / sec to 4.79 Ian / sec at the age of504 days.
Under prolonged air curing all the HVF A concretes behaved somewhat
similarly, and showed a pattern of behaviour similar to that of dynamic modulus of
elasticity - a significant fall with age soon after the cessation of water curing, even
though the concretes had high strength. The results show that mixes 350S and 450S
performed best followed by those incorporating both fly ash and silica fume. At the
end of 504 days, the pulse velocity development of HVF A concretes ranged from
3.98 Ian / sec to 4.13 Ian / sec.
111
Chapter 4 Test Results and Discussion of Part I
112
5 r-----------------------------------____~
350F 350SF 350S 450F 450SF 450S
o o
•
4.75
u
(\)
~ 4.5
]
~.
'0 4.25
.....o
~
(\)
-3
~
4
3.75
Figure 4-39 Variation of pulse velocity on HVFA concrete under air curing
112
Chapter 4 Test Results and Discussion ofPart I 113
The results indicate that exposure to ambient air of all HVF A concretes
caused a slight retrogression of ultrasonic pulse velocity with continued air drying.
The pulse velocity of air cured concrete is about 85 % of the average velocity of
water cured concrete at 504 days. The percentage gain of pulse velocity with
continued air drying is almost the same as the compressive strength development
under similar conditions. In general, the more dense and strong the HVF A concrete,
the lower is the value of the reduction. At the end of 504 days, the values of the pulse
velocity were in a manner similar to those of dynamic modulus of elasticity. The
average percentage loss on pulse velocity for all HVF A concretes is somewhat lower,
about 3 %, compared to the same concretes at the age of 28 days. The reduction in
pulse velocity is almost the same as the loss in dynamic modulus of elasticity, as was
also found by others [41].
From the preceding data in Tables 4-23 and 4-24, similar rates of
development of dynamic modulus of elasticity due to total binder content were
observed. Therefore, for a given aggregate and a given richness of the mix, the pulse
velocity of the concrete is affected more by the modulus of elasticity, and
consequently, no unique relation between pulse velocity and compressive strength
exists [78].
113
Chapter 4 Test Results and Discussion of Part 1 114
5 5
-- - "_OM' -
~ 4.75 V ~ 4.75
] 4.5 ] 4.5
.~ 4.25 .~4.25
() ()
.Q 4 .Q 4
~ ~
~
~ 3.75
3.5 13~.~45OC I -~
~ 3.75
3.5
3.25 3.25
0 112 224 336 448 560 0 112 224 336 448 560
Days Days
5 5
~ 4.75 ~ 4.75
Figure 4-40 Effect of total binder content on pulse velocity ( Water curing)
5 5
~ 4.75
V
~.. .-................. --- - ... ~ 4.75
'" '"
] 4.5 ] 4.5
.~4.25 .~ 4.25
() ()
.Q 4 .Q
~ ~
~
~ 3.75
3.5 13~~C450C I -~ ~ 3.75
3.5
3.25 3.25
0 112 224 336 448 560 0 112 224 336 448 560
Days Days
5 5
~ 4.75 ~ 4.75
] 4.5 ] 4.5
~
.~ 4.25 ~ .~ 4.25 Ir' -
-~
() ~
~'
()
0 v v 0
4
4
~
'V
~
-~
~ 3.75
3.5 17 13:F45~F I ~
3.75
3.5 r 3!S4~SI
3.25 3.25
0 112 224 336 448 560 0 112 224 336 448 560
Days Days
Figure 4-41 Effect of total binder content on pulse velocity (Air curing)
11-'
Chapter 4 Test Results and Discussion o{Part 1 lI5
Table 4-30 represents some typical data of ultrasonic pulse velocity for the OPC
concrete mixes under both types curing conditions. The rate of gain of pulse velocity
as a % of 28 day velocity are is shown in Tables 4-31 and 4-32.
From the results, it can be seen that, under continued moist curing condition,
the percentage increase in pulse velocity from 28 days to one year is 3 % to 4 % for
all of normal OPC concretes. The corresponding average increase for the HVFA
concretes was 5 % to 13 %. The differences of between 350F and 350C, 450F and
450C mixes in pulse velocity at one year under water curing were about 10 % and 7
%, respectively.
At the end of one year, very little reduction in pulse velocity of OPC concrete
with air curing was observed. The differences between 350F and 350C, 450F and
450C mixes in pulse velocity at the end of on year under air curing were only 4 %
and 2 % respectively. Undoubtedly, the potential for pulse velocity development at
later ages for HVFA concretes exists, even if at a slow rate.
115
Chapter 4 Test Results and Discussion o[Part 1 116
Table 4-31 Rate of gain velocity as a 010 of 28 day pulse velocity ( Water curing)
(Unit: % )
Table 4-32 Rate of gain velocity as a % of 28 day pulse velocity ( Air curing)
(Unit: % )
The relationship between the pulse velocity, V, and the dynamic modulus of
elasticity, ED, can be represented by the following equation:
116
Chapter 4 Test Results and Discussion o(Part I 117
v = a + b ED ------------------ ( 4-19 )
The binder content had shown similar influence on pulse velocity under both
curing conditions, the total binder content had a marked influence on the dynamic
modulus of elasticity. Further, the wet concretes had generally higher values of pulse
velocity and dynamic modulus of elasticity than those under dry curing. It is therefore
suggested that separate equation be used for each curing condition and total binder
content.
The variation of pulse velocity with dynamic modulus of elasticity for HVFA
concretes under the water curing is shown in Fig 4-42.
5.2 ~------------------------------~------------~
4.8
(.)
(l)
~
S
...!:oI::
4.6
..-
~ 4.4 •• •
.....
(.) o ••
o
'"a) 4.2
o
• ••
~ o
o
••
(l)
-S
~
4
o
3.8
3 3
350 kg/ m 450 kg/ m
3.6 o • Water curing
3.4 ~~~~~~~~~~~~~_L~~~~~~~~~~
30 35 40 45 50 55 60 65
Dynamic modulus of elasticity, GPa
The relationship is based on measurements from 1 day to 504 days for moist
cured HVFA concretes. Regression analysis of the data shown in the figure produced
117
Chapter 4 Test Results and Discussion of Part I 1J8
v = 0.86 + 0.45 ED 3
( 350 kg/m mixes) (r = 0.99 ) ------ ( 4-20 )
v = 0.84 + 0.44 ED 3
( 450 kg/m mixes) (r = 0.99) ------ ( 4-21 )
For the wet cured concrete, the two equations above are more or less
identical the same for the two total binder content concretes. The linear regression
shown in the figure has a correlation coefficient of 0.99. These data indicate a
significant correlation between ultrasonic pulse velocity and dynamic modulus of
elasticity for HVFA concretes at the 99 % confidence level.
In Fig 4-43, results of the one day to 504 days pulse velocity are plotted
against the dynamic modulus of elasticity for the air cured HVF A concretes.
Regression analysis of the test data shown in the figure produced the following
equations:
5.2 .----------------------~
~ 4.6
]
~ 4.4
.....
u
o
Q) 4.2
~
(\)
~ 4
~
3.6
3.4 L-..l.-"/::::::..~--L....//___'-_L-__L_---L---L-J.-L---'-~__'_L___L__'___~__'_~_____'
32 34 36 38 40 42 44 46 48 50
Dynamic modulus of elasticity, GPa
118
Chapter 4 Test Results and Discussion olPart 1 119
Again, the correlation factor between the two sets of equations is still high,
and the equation relating the two is approximately the same for the two total
cementitious content concretes. Irrespective of the curing condition, the pulse
velocity is controlled primarily by the dynamic modulus of elasticity of the concrete,
and ultrasonic pulse velocity increases with an increase in dynamic modulus of
elasticity.
It is worth pointing out, from the data for both types of curing conditions,
that concretes with the lower binder content have a marginally higher pulse velocity
than that with the higher binder content. This is in agreement with the concept, in
general, that the pulse velocity varies inversely to the density of concrete- V2 = ED / p,
where p is density.
The effect of the variables such as, mix proportions, and age of HVF A
concretes on carbonation were studied in a programme over 18 months under both
curing conditions. The relationship between carbonation depth and compressive
strength was also investigated.
119
Chapter 4 Test Results and Discussion of Part 1 120
The results of this investigation on carbonation for all the concrete mixtures under
both curing conditions are shown in Tables 4-33 and 4-34. For purposes of
comparison, the results of the normal OPC concrete on carbonation depth are also
included in the above table. Figs 4-44, 4-45, and 4-46 show the relationship between
the exposure period and carbonation depth of various HVFA and OPC concretes
stored in air. Figs 4-47 and 4-48 show typical examples of carbonation ofHVF A and
OPC concretes stored in air at various ages.
180 Min. 0.0 1.0 0.0 6.0 0.0 4.0 0.0 3.2
360 Min. 0.0 1.5 0.0 9.0 0.0 6.0 0.0 5.0
540 Min. --- --- 0.0 11.2 0.0 10.5 0.0 10.1
120
Chapter 4 Test Results and Discussion of Part I 121
16
350C 350F 350SF 350S
14 o. X 0 0
12
§
tCl)
10
Cl
d 8
.....0
~
d
0 6
~
U
4
2 o
0'
14 28 90 180 360 540
Age (Days)
Figure 4-44 Carbonation depth of concrete - 350 kglm3 mixes ( Air curing)
At one year, HVF A concretes that were water cured initially for 7 days
showed a significantly higher depth of carbonation than the corresponding normal
ope concretes for both types of binder contents. The depth of carbonation in normal
ope concrete is, on average, only 25 % of that of the HVFA concrete. The higher
carbonation of HVF A concrete is attributed to greater permeability of the surface
concrete containing fly ash.
121
Chapter 4 Test Results and Discussion o[Part I 122
180 Min. 0.0 1.7 0.0 4.0 0.0 3.5 0.0 2.8
360 Min. 0.0 2.0 0.0 4.0 0.0 3.5 0.0 5.0
540 Min. --- --- 0.0 11.0 0.0 11.1 0.0 11.0
122
Chapter 4 Test Results and Discussion ofPart 1 123
16
450C 450F 450SF 450S
14 D I •
12
~
t Q)
10
0
d 8
.....0
~
d
0 6
~
u
4
0 1
16 16
~ 14 I 350C 450C I ~ 14
.g, 12 .g' 12
8'10 8'10
§ 8 § 8
.~
6 ".g 6
=
0
4
=
0
..D 4
~ t;J
u 2 u 2
0 ~ 0
14 28 90 180 360 14 28 90 180 540
Days Days
16 16
~ 14 350SF 450SF
-<7 ---
§14
.g' 12 ..d' 12
8'10 Q..1O
8
§ 8
"+=I
=
0
8
.~
~
t::
6 6
t::
0 0
... 4
..0
~
..0
t;J
4
u 2 u 2
0 0
14 28 90 180 540 14 28 90 180 540
Days Days
Figure 4-46 Effect of total binder content on carbonation depth ( Air curing)
123
Chapter 4 Test Results and Discussion of Part 1 12-1
,
'"
6-
I
... • •l
"• -
.t/ "
,
450f -I RO
50I ,- IX i\!
12-l
Chapter 4 Test Results and Discussion of Part I 125
Further, under air curing, it is also clear from the above data that concrete,
Mix-350F and Mix-450F, exhibited higher carbonation depths than those obtained in
the concretes incorporating both fly ash and silica fume.
At the end of 18 months, for all HVFA concretes under air drying, the depth
of carbonation was a great deal higher than the values received at one year. For
specimens stored in air, differences in carbonation depth were found for concrete
mixes 350F and 450F, and the mixes containing fly ash as a partial replacement for
both sand and cement.
The influence of fly ash and silica fume on carbonation of concretes found in
this study is in agreement with the results of other investigators [27,44,79]. The
increase in the carbonation depth of HVFA concrete compared to ope concrete may
be due to the individual or combined effect of their greater surface porosity, and its
smaller concentration of carbonatable constituents [5]. This indicates that the
addition of such mineral admixtures to concrete which does not form proper matrix
films upon coalescence, may rather increase the risk of carbonation. Kokubu &
Nagataki also observed that the carbonation was distinctly increased due to raising
the fly ash replacement ratio, among other things [80].
The important influence of curing condition on carbonation is clearly seen in
those results. As for depth of carbonation, the depth increases the longer the period
of exposure in air, which is considerably higher than that of the concrete exposed to
moist curing. The results are also relevant for the strength development as well as
dynamic modulus of elasticity, ultrasonic pulse velocity for high strength concretes. If
the concrete matrix is porous due to carbonation, strength reduction generally
occurs. As discussed in preceding sections, strength development of HVF A concrete
in case of curing in water increases with age. When cured in air, although there is an
increase in strength seen from the age of four weeks to one year, a trend of decline is
recognised. Similar trends were observed for the dynamic modulus of elasticity and
pulse velocity. These results indicate that high strength alone is not sufficient to
prevent carbonation, moisture is also needed to slow down the carbonation process.
These findings are in accordance with the results of other investigators [48,81,82,83].
Many of them suggest that an increase in curing time of the concretes with admixture
125
Chapter 4 Test Results and Discussion olPart 1 126
reduces the difference with ope concrete regarding the rate of carbonation, as well
as other physical and mechanical properties of concrete.
e =a +K Ji ---------------- ( 4-24 )
Mix a K r
126
Chapter 4 Test Results and Discussion o{Part 1 127
16
350C 350F 350SF 350S
14 0 )( 0 0
12
§
~ 10
x
fr
0
§ 8
.+=!
Cd
§ 6 o
x
~
u
4 o
2 o o o
0
1,000 2,000 3,000 4,000 5,000 6,000
Time, Sec
16
450C 450F 450SF 450S
14 o •
12
~
o£3" 10
fr
0
§ 8
.+=!
Cd
§ 6
~ + + •
u
4
2 o
127
Chapter 4 Test Results and Discussion ofPart 1 128
These parameters are valid for exposure times up to one year. All the HVF A
concretes show higher carbonation rate coefficient, K, in comparison to normal ope
concretes because of the reasons mentioned earlier. Further, concrete with the 350
3
kg/m binder content had a higher carbonation rate coefficient than that of the 450
3
kg/m mixes. Again, the concretes made of 350F and 450F exhibited the poorest
results.
In general, the behaviour of concrete is always based on the assumption that the
ambient medium, i. e., air does not react with the hydrated cement. In fact, structural
concrete is always exposed to the atmosphere which contains high levels of carbon
dioxide. In the presence of moisture, the carbon dioxide reacts with the hydrated
cement, and this results in carbonation taking place through the pore system in the
hardened cement concrete, which may result in a decrease in strength. The rate of
carbonation depends on the atmospheric carbon dioxide, cement content, relative
humidity, temperature, and the porosity of the concrete, the latter depending upon
the type of cement, the water / binder ratio, and the degree of hydration.
Regression analysis of the data shown in the figure produced the following
relationships:
where fc is the compressive strength in MPa, and e is the carbonation depth in mm.
128
Chapter 4 Test Results and Discussion ofPart I 129
It is often said that the rate of carbonation is simply a function of the strength
of concrete. As shown in the figure, there is a reasonable inverse correlation between
the depth of carbonation, C, and compressive strength, fe, of HVF A concrete : in
other words, the depth of carbonation of concrete increases with decrease in strength.
As for depth of carbonation, the depth increases the longer period the of time
left in air. At the end of 540 days, although the carbonation and strength relationship
shows a lower correlation coefficient, the data indicate still significant correlation at
the 99 % confidence level between the depth of carbonation and compressive
strength. On the basis of the present studies, of the factors affecting carbonation,
strength appears to be an important factor.
90 ~----~--------------------------------------,
80
0
c;I
p. 180day
:E~ 70 Air
t
~
rJJ 60
+ + 0
0
360day
Air
.-
Q)
>
t il
til
Q)
+
+ 0
0
0
0
540day
+
E' 50
'""' Air
0
U
0
0
8
40
Carbonation Depth, mm
129
Chapter 4 Test Results and Discussion of Part 1 130
130
Chapter 4 Test Results and Discussion ofPart 1 131
4-9. CONCLUSIONS
This chapter has examined the effect of curing condition on the engmeenng
properties of HVFA concrete with and without silica fume. The following
conclusions are based on the results and analyses presented in this chapter.
1. For the HVFA concrete investigated, continuous moist curing gave higher
strength, dynamic modulus of elasticity, and ultrasonic pulse velocity than
prolonged exposure to air drying.
3. Air curing preceded by 7 days moist curing is beneficial to the flexural strength
development of HVFA concretes. At 90 days, concrete under this condition
exhibited the highest flexural strength.
3
5. In general, increasing the binder content of the concrete from 350 kg/m to 450
kg/m3 of binder content had little effect on strength, but did yield a large benefit
in terms of dynamic modulus of elasticity, and pulse velocity in most of the
HVF A concretes.
6. Under the same conditions, concretes made with both fly ash and silica fume had
engineering properties in most respects which were as good as, or superior to,
those made by fly ash alone as cement replacement. The use of fly ash to replace
both cement and sand was able to combine the benefits of both separate
replacements.
131
Chapter 4 Test Results and Discussion of Part 1 132
of dynamic modulus of elasticity, or pulse velocity for all HVF A concretes was,
on the other hand, very low, only about 3 %, when compared to the same
concretes air cured for 28 days.
8. The deterioration of concrete also influences the cement paste in the interior of
HVF A concrete, and has a close relationship with carbonation. The phenomenon
of a decline in engineering properties accompanying increase in depth of
carbonation is also seen in normal portland cement concrete.
9. Although the early strength of concrete made with HVFA was found to be low,
those at ages beyond 28 days showed a continued rate of development of
hydration, superior to that of the normal concrete. The HVFA concrete in this
study has good potential for use in structural and mass concrete applications.
10. The swelling in the HVFA concretes was associated with formation of ettringite.
The ettringite formed in the early ages of hydration was more efficient in
producing swelling. This suggests that swelling began at a lower critical degree
of hydration ; the slower rate hydration of the HVFA concrete probably resulted
in slightly more swelling due to adsorption of water.
11. Drying shrinkage did not correlate with total porosity but correlated reasonably
well with the volume of micropores plus mesopores.
12. The drying shrinkage strain measured on all HVFA concretes was comparable to
Ross's formula, and was about 95 % of the average percentage of the latter.
3
13. The values of static modulus of elasticity, Be, of 450 kg/m mixes under water
curing are higher than the calculated values according to ACI building Code
318-89.
14. The dynamic modulus of elasticity varies with changes in the total binder content
similarly to normal portland concrete mixes, and is curvilinear in its relation to
compressive strength.
132
Chapter 4 Test Results and Discussion of Part I 133
15. There is no physical relation between the ultrasonic pulse velocity and strength.
The pulse velocity was found to be very sensitive and related to the density of
concrete, and linearly related to the dynamic modulus of elasticity. The total
binder content does not seem to have much effect on pulse velocity under both
types curing conditions. Under moist curing, all of HVF A mixtures appeared to
have more or less the same value of pulse velocity beyond about six months.
16. Overall, the depth of carbonation was found to be related to the compressive
strength of concrete, with and without fly ash. Generally, the HVF A concrete
carbonated to somewhat higher depth compared to the normal portland concrete
exposed to air curing.
17. High strength alone is not sufficient to prevent carbonation, moisture is also
needed to slow down the carbonation.
133
Chapter 5 Microstructure of Hydrated Cement Pastes 13.+
CHAPTER FIVE
5-1. INTRODUCTION
There is, thus, a great impetus to study the microstructure of such a concrete
in order to examine the changes of hydration of HYF A concrete progressing with
time. To a lesser extent, the influence of the characteristics of microstructure are also
discussed but, it is useful to review the pore structure and mineralogical compositions
of concrete.
The pores formed in fresh concrete are either water-filled or gas-filled. At the initial
mixing period, the cement grains, and the conglomerates in the fresh pastes, are
separated individually by the mixing water, which forms the water-filled space. The
134
Chapter 5 Microstructure of Hydrated Cement Pastes 135
water-filled space in the freshly mixed cement paste represents space that is available
for the formation of cement hydration products.
As hydration proceeds, the volume of this space, which has a bulk volume
larger than that of the original unhydrated cement, is continually reduced by the
formation of the hydrate products that form a continuous matrix and bind the residual
cement grains together over a period of time. At any time, that part of the original
water-filled space not occupied by hydration products constitutes part of the pore
system of the paste, and generally speaking, that part with the largest pores sizes.
Hydration reduces the size and volume of this capillary space. If the original capillary
space is small, the bulk volume of the gel eventually will be sufficient to fill this space
and produce a paste free from capillary space.
In general, the volume of the water filled spaces in fresh paste must exceed
the absolute volume of the cement, or some of the original cement must remain
unhydrated. For this reason, the hydration products are not sufficient to fill the
original water space in the paste, even after complete cement hydration, and pastes
having a relatively large volume of capillaries will be produced.
Studies on the hardened cement pastes have shown that the size of its largest
pores depends principally on the degree of hydration, varying from about 111m for
young pastes in which the cement hydration has progressed only a little, to about
O.ll1m in mature pastes that are more completely hydrated.
135
Chapter 5 Microstructure of Hydrated Cement Pastes 136
has such a low water / cementitious ratio that not all the cement can even become
hydrated, the gel will fill completely all the space not occupied by unhydrated cement,
and will contain porosity.
Pore sizes have been classified in different ways by Powers, Feldman and
Sereda, Diamond et al , IUAPC, and Bentur et al [85,86,87,88]. A comparison of
pore classification is shown in Fig 5-1. Table 5-1 shows a classification of porosity in
cement paste by Powers model, it can be seen that there is an enormous range of pore
sizes, from 10 pm to less than 0.005 pm ( 0.5 nm) in diameter.
,----
Reference
Diamond
I Interlayer
I
136
Chapter 5 Microstructure of Hydrated Cement Pastes 137
All porous materials have this same inverse dependency between strength and
porosity. Both theory and experiment lead to relationships that are either parabolic or
exponential between the two. The effect of porosity on the strength of hydrated
cement paste has been studied widely. Many investigators [89] have recognised that
porosity is an important parameter in determining strength of hardened cement
pastes, and its effect is expressed by a power function of the type :
fc = fo xn --------------- ( 5-1 )
l37
Chapter 5 Microstructure of Hydrated Cement Pastes 138
0.68a
X = --------------- ( 5-2 )
0.32a+%
This knowledge has been used in preparing cement paste; cement is mixed
with only enough water to make it workable, thereby minimising the water remaining
in pores after the paste hardens.
Relationship between the strength and gel/space ratio is shown in Fig 5-2.
The relationship is based on results tested from 2 in. mortar cubes [69]. A linear
relation between strength and porosity, within the range of the latter between 5 and
28 %, was established by Rossler and adler [90].
Watson [91] has pointed out the importance of porosity and derived a
formula which expresses the relationship between strength and V s / Vp , the solid /
pore volume ratio, where V s and Vp are the volume fractions of the solid material and
pores respectively as determined by helium pycnometry. Meanwhile, Feldman and
Beaudoin [92] have emphasized that the porosity is a major factor in controlling
strength, and modulus of elasticity of normal OPC concrete systems. The relation
between the strength of mortar based on volume of pores larger than 20 nm in
diameter is shown in Fig 5-3 [93].
138
Chapter 5 lvficrostructure o{Hydrated Cement Pastes 139
140
&! 120 •
~
rA'
.B::s 100
U
-.
.-
d 80
--~
C"'I
60
.......
V)
~
0
40
t -t:.
83
-=
rF.l 20 •
o L -_ _~_ _ _ _~_ _ _ _~_ _ _ _~_ _ _ _~_ _~L- __ ~ ____ ~
Figure 5-2 Relation between the compressive strength of mortar and gel/space
ratio [69]
60.---------------------------------------------
50
• • •
•
• •
9 10 11 12 13 14 15
Porosity-per cent
139
Chapter 5 JvIicrostructure of Hydrated Cement Pastes 1-10
From above analysis and further studies, Young and co-workers [96]
conclude that shrinkage is the result of two components which are superimposed
upon one another :
1. Capillary drying shrinkage - loss of moisture from the "pore component", meso-
pores: 25 A ~ 300 A, which is assumed to be due to capillary effects; and
2. Gel drying shrinkage - loss of moisture from the "C-S-H component" which
includes the C-S-H particles and the pores smaller than 25 A (micropores ).
Establishing the pathways and mechanisms for fluid flow through cementitious
materials is an important part of the development of the understanding of mechanisms
of ionic transport. The permeability is affected by many factors such as hydrostatic
pressure, osmotic effects, pore differentials, and temperature differentials. The
140
Chapter 5 Microstructure o(Hydrated Cement Pastes ].11
dv h
- = K - --------------- (5-3 )
dt x
In a fresh paste, the flow of water is controlled by the size, shape, and
concentration of the original cement particles. With the progress of hydration, the
permeability decreases rapidly because the gross volume of gel ( including the gel
pores) is approximately 2.1 times the volume of the unhydrated cement, so that the
gel gradually fills some of the original water-filled space. As hydration proceeds, the
capillary network is cut offby C-S-H gel, and then the capillary pore system becomes
discontinuous. Flow is now controlled by micropores and this is very slow. This is
accompanied by a continuous decrease in coefficient of permeability, "K".
In general, water can flow more easily through the capillary pores than
through the much smaller gel pores. The cement paste as a whole is 20 to 100 times
more permeable than the gel itself [69]. In a mature paste, the permeability depends
on size, shape, and concentration of the gel particles, and on whether or not the
capillaries have become discontinuous. It is worth to note that low-pressure steam
curing shows a slight increase in permeability [97].
If the specimen is not cured well, drying shrinkage may rupture some of the
gel between the capillaries, and thus result in larger permeability coefficient values.
On the contrary, the specimen cured sufficiently in water has more moisture, and it
has denser structure because of more hydration than in specimen with drying.
Goto and Roy [94] presented some experimental results showing the
existence of a relationship between permeability, and pore size distribution. The
relationship between porosity and permeability was also demonstrated by Powers
141
Chapter 5 Microstructure of Hydrated Cement Pastes 142
who confirmed that the coefficient of permeability decreases with a decrease in the
percentage of capillary porosity.
The term "absorption" is usually applied to concrete in regard to the weight gain of
partially dried specimens upon contact with or immersion in water. However, water
can enter the pores, the amount of the absorbed water depending primarily on the
abundance and continuity of the pores in the particle, whereas the rate of absorption
depends on the size and continuity of these pores.
During absorption, it may be considered that the larger capillary spaces in the
paste are the first to be wetted, with the finer gel pores perhaps next, their rate of
absorption being slow because of their low water permeability. Thus, these features
of the pores are more important than merely their total volumes as reflected by the
magnitude of water absorption.
142
Chapter 5 Microstructure ofHydrated Cement Pastes 143
It may be noted that gravel has generally a higher water absorption than
crushed rock of the same petrological character, because weathering results in the
outer layer of the gravel particles being more porous and absorbent [69].
From further studies, it is generally held that the water absorption of mature
concrete is not a simple function of its porosity, but depends also on the pore size,
distribution, water / cementitious ratio, aggregate characteristics, air content, cement
type and fineness ( particularly at early ages ), specimen size and shape, method of
surface preparation ( cast, broken ), degree of hydration, continuity of the pores, and
so on.
Research into the hydration of cement has been directed, on the one hand, towards
the phenomenon of engineering characteristics in term of microstructure and
porosity, and, on the other, to the way in which the chemical composition of the
cement and its hydration products can affect its properties or be modified to improve
them.
143
Chapter 5 Microstructure ofHydrated Cement Pastes ]./-1
Chapter two briefly dealt with the hydration reaction of portland cement in
general, and cements differing in chemical composition may exhibit different
hydration products when hydrated.
Upon addition of water the cement begins a series of hydration reactions, the
most important of which is the formation of a poorly crystallised hydration product
called calcium silicate hydrate gel ( C-S-H gel ), and calcium hydroxide ( CH ) by the
calcium silicates. Other hydration products involving iron, aluminium and sulphur
constituents are also formed. Thus, this section deals with the hydration products of
portland cement, and its crystal structure, and properties relevant to the present aim
of the investigation.
The C-S-H gel formed in cement paste has been thought to have a highly disordered
structure with a small particle size ( dimension small than 150 A ), and a high surface
area [101]. It is termed "gel" because its dimensions are in the colloid range ( less
than 1000 A ) and is poorly crystalline, being essentially amorphous. Jenning and
Pratt [85,102] have shown the thin sheets or foil-shaped C-S-H structures in the
saturated state from T .E.M. ( Transmission Electron Microscope) observations.
The similarity between C-S-H gels and tobermorite has been demonstrated by
thermal analysis, electron and X-ray diffraction, and infrared spectroscopy. Results of
X-ray diffraction analysis, however, remain the main basis for this association
between C-S-H gels and tobermorite. The C-S-H gels give three broad lines
corresponding to 3.05 A, 2.8 A, and 1.82 A reflections of crystalline tobermorites
[103].
H
I **** IGeI in Paste
CSHI
9T
SL-----------~~----~~~------~c
C/ s= 1.3 C/ s= 2.3
L_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ ~
The chemical compositions of C-S-H gel are quite diverse, especially when
other chemical compounds are added, and it is believed that solid state substitutions
will occur, which has been confirmed by electron beam micro-analysis [104].
1~5
Chapter 5 Microstructure of Hydrated Cement Pastes 146
The physical characteristics of C-S-H are also diverse, and these characters
dependent on the technique of measurement, the sample preparation procedure and
the environmental condition. The mechanical properties of the hydrated gel are
described by Powers and Brunauer, who presumed that the particles are to be held
together by Van der Waals forces. Swelling on exposure to water is explained by the
individual particles separating due to layers of water molecules existing between
them, and this model also includes the existence of some chemical bonds between the
particles to explain the limited swelling nature of the materials. Feldman and Sereda
have postulated that the interlayer regions of C-S-H play an important role in
shrinkage, particularly irreversible shrinkage.
The exact nature of the C-S-H microstructure is still uncertain, due mainly to
the lack of direct structure determination techniques [108]. However, the C-S-H gel
is still considered to be primarily responsible for the strength and cementing
characteristics of the concrete. Also, it is now believed that the removal of water by
any drying process will more or less change the C-S-H structure, and that the dried
paste will be different from the saturated paste. The various morphologies and the
various compositions of C-S-H also cause problems in understanding.
146
Chapter 5 Microstructure of Hydrated Cement Pastes 147
crystals, thin plates, elongated crystals, flakes, or variation of these forms [87]. The
phase of calcium hydroxide gives six strongest X-ray diffraction peaks corresponding
to the 2.63 A., 4.90 A., 1.93 A., 1.80 A., 3.11 A., and 1.69 A. Its morphology is
particularly affected by admixtures and by the temperature of hydration.
The ettringite takes the form of needles, and because of its relatively low
specific gravity ( 1.70 ) occupies a correspondingly large volume. Excessive
formation of ettringite may lead to tensile stresses in the concrete, which may lead to
cracking of the concrete.
147
Chapter 5 Microstructure o(Hydrated Cement Pastes U8
occurs sporadically, and to a limited extent in portland cement concrete, and is more
commonly found in lime-pozzolan systems.
5-7-4. Carbonates
148
Chapter 6 Experimental Programme and Characteristics oU..,1aterials in Part II 149
CHAPTER SIX
6-1. INTRODUCTION
During this study, mercury intrusion porosimetry (MIP ) was introduced and
used to perform pore size distribution analysis to characterize the particle size of the
cementitious constituents of different mix proportions. X-ray diffraction ( XRD )
analysis technique is used to identify the mineralogical products, and monitor the
progress of hydration. This study also examines the effects of HVF A concrete on
permeability and water absorption, in order to investigate the density of HVF A
concrete under pozzolanic activity, and the effect of short and long term hydration
reaction. In this chapter, the experimental programme and the preparation of test
samples used in Part II of the research project are described.
149
Chapter 6 Experimental Programme and Characteristics ofMaterials in Part 11 150
The material used and mixing proportions of the mixes were the same as those used
in Part I of the project, and have been discussed in Chapter 3, Section 3-3.
In the present investigation, a series of concrete samples of the same concrete mix
proportions were tested at corresponding ages to correlate microstructure
development with strength characteristics. Immediately following compressive
strength testing of the specimens, experimental samples were taken from the centre of
each specimen after cutting away all surfaces to be free of the mold-wall impression.
The samples were then split into smaller pieces and immersed in high purity methanol
in order to avoid further hydration of cement, and dissolution of hydration products.
The samples were kept in methanol for at least 21 days before use. The
microstructure mechanism tests were carried out for all the concretes after 28, 90
days and 18 months of moist and air curing.
The technique used to measure the pore size distribution was mercury intrusion
porosimetry. Mercury intrusion porosimetry is based on the capillary law governing
liquid penetration into small pores. The pressure required intrude the pore is a
function of the contact angle, the surface tension of the intruding liquid, and the
shape of the pore to be intruded.
150
Chapter 6 Experimental Programme and Characteristics ofMaterials in Part II J5J
y<l> cose
D=- --------------- ( 6-1 )
P
where:
e = the contact angle between the mercury and the pore wall
<I> = the shape factor of the pore
for the cylindrical and the slit pore is 4 and 2 respectively, while the elliptical pores
take the values between 2 and 4 [109].
The contact angle ( e ) between mercury and the solid containing the pores
varies somewhat with the methods adopted for drying the concrete sample, maturity
of the concrete and solid composition [110]. Winslow and Diamond [111] measured
the contact angles ( e ) for specimens oven dried at 105°C in oven and dried over
magnesium perchlorate hydrate. These measurements indicated contact angles ( e )
of 117 degrees and 130 degrees for the oven dried samples and perchlorate dried
respectively.
151
Chapler 6 Experimental Programme and Characteristics oOlaterials in Part 11
The pore size distribution studies were carried out using a Micromeritics
Poresizer ( Model 9320 ). A photograph of the instrument is shown in Fig 6- 1
Samples used for the porosity test were fragments of concrete by taking from the
internal parts of specimens with all coarse aggregate removed. These fragments were
then broken into pieces of approximately 3 rnm, and immersed in methanol for at
least 21 days. Prior to testing, the samples were oven-dried until they were of a
constant weight.
About 1.5 - 2 grams of sample were then weighed, and inserted into a glass
penetrometer tube of known weight and volume. This penetrometer tube was then
fitted with a sealing cap, and a special grease ( Apiezon H ) applied to provide an aIr-
tight seal. The sealed penetrometer tube was then inserted into the low-pressure port
152
Chapter 6 Experimental Programme and Characteristics ofMaterials in Part II 153
After de-gassing, the penetrometer tube was filled with mercury, at a filling
pressure of 0.5 - l.5 psia, and the combined weight of the penetrometer, sample and
mercury was recorded. The penetrometer tube was then placed in the high pressure
port of the poresizer, where pressure was increased slowly from atmospheric up to
30,000 psia, and the amount of mercury intrusion measured at various points during
the pressure increase. A dedicated computer fitted with a maths co-processor was
used to control the instrument to record data, and to calculate various parameters.
The penetrometer used had a bulb volume of 5.89 mL and a stem volume of
1.131 mL. The value used for the surface tension of the mercury ( y ) was 485.00
dynes / cm. The shape factor ( <I> ) used was equal to 4, assuming all the pores to be
of cylindrical shape. For the pore diameter calculations a contact angle ( e ) of 117
degrees was used for the present samples, ( oven dried to 105°C to a constant
weight ), based on the previous investigation carried out by Winslow and Diamond.
Table 6-1 presents a typical output obtained with a pressure run from the present
study.
153
Chapter 6 Experimental Programme and Characteristics ofMaterials in Part II 154
LOW PRESSURE :
MERCURY FILLING PRESSURE: 1.4230 psia
LAST LOW PRESSURE POINT: 1.4230 psia
HIGH PRESSURE :
RUN TYPE : AUTOMATIC
RUN METHOD: EQUILIBRATED
EQUILIBRATION TIME: 5 seconds
MAXIMUM INTRUSION VOLUME: 0.0010 mL/g
15.t
Chapter 6 Experimental Programme and Characteristics ofJvfaterials in Part II 155
155
Chapter 6 Experimental Programme and Characteristics ofMaterials in Part II 156
156
Chapter 6 l'.:xperimenfal Programme and Characteristics of.\Jareria/s in Pan 1I 157
At the specific age of the air permeability test, a set of three specimens were drilled
from the full thickness of the concrete to produce 50 mm diameter core samples
These cylindrical cores were then sliced at each end to produce samples of between
30 ~ 40 mm height for air permeability testing. The test apparatus was developed at
the University of Leeds [112] and a general view of the cells in use is shown in Fig 6-
2.
The permeability cell consists of a sample holder, pressure gauge, gas supply
and flow meter at the downstream side. The cell was designed to test concrete
specimens of 50 mm diameter and 10 mm to 50 mm height. Due to the sensitivity of
oxygen permeability measurements to the presence of moisture content in the
concrete, after the prescribed period of curing, the specimens were oven dried at 105
°C to constant weight, until weight change was less than 0.1 percent over a 24 hour
period ( BS 1881 : part 5 ). The sample was placed in a rubber cylinder inside a steel
ring cylinder. Gas was forced to flow in the vertical direction through the sample
157
Chapter 6 Experimental Programme and Characteristics ofMaterials in Part II 158
Flow rates were recorded by using a 1.7 mm diameter bubble flow meter at one bar
pressure when a steady state of flow occurred.
The water absorption test in the present study was according to BS 1881 : Part 122 :
1983 "Method for determination of water absorption". Similarly, a set of three
specimens were core-drilled from the full thickness of the concrete to produce
cylindrical samples of75 mm diameter, with a height of 100 mm. These samples were
then used for determining water absorption. The core samples were then oven dried
at 105°C for 72 + 2 h, removed from the oven, and cooled for 24 ± 0.5 h in a dry
airtight vessel. Each sample was weighed and immediately and completely immersed
in a tank with its longitudinal axis horizontal and at a depth such that there was 25 ±
5 mm of water over the top of the sample. The samples were kept immersed in the
water tank for 30 ± 0.5 min. Each sample was removed and shaken to remove the
bulk of the water and dried with a cloth until all free water was removed from the
surface. Each sample was then re-weighed.
The measured water absorption, Woos, of the sample was calculated as the
w abs
= WSSD -
W
WOD X 100% --------------- ( 6-2 )
OD
where WSSD and WOD represent the weight of the concrete sample in the SSD (
158
Chapter 6 Experimental Programme and Characteristics o[JJateria/s 111 Part 11 159
1.2 r-----:----:-----:--~---------------,
~
1 , . . ,. . . . ;; ·r·r······ t------------:------------I------------1---
o ~ [
'0
~ 0.9 ---------t----- ! i : i . .
u
§
.+=1
o
u
~ 0.8 1: ... 1<::·1:::"; f
0.7 ····,I····L····J,4
. , , . , .
0.6 ---------~ ---_.. ----1 .. -. ------i- --------- -~ --~.
+ _. - -~ --~
- - -- -- - - --- --i. + •• - _ •• - --. L -- --_. - ---~ ------ ------~ ----------- -~- -------.. -- ~---.
~ . . . ~ ~~.
O. 5 L--L....-l---'-----L----'---'-----'------L---L--L--'---..L.-.1-...-JL-..!--1.-L--L----'-~__"____L_...L..._...l
!
30 40 50 60 70 80 90 100 llO 120 130 140 150
Length in mm of 75 mm dia. specimen
Figure 6-3 Correction factor related to the length of the specimen [113]
X-ray radiation has wavelengths in the same range and is scattered by atoms
through an interaction with the electrons. The diffraction phenomenon can be
visualised as a reflection by a set of parallel planes of atoms lying in the plane.
Diffraction occurs when the Bragg Equation [114] is satisfied:
Were:
159
Chapter 6 Experimental Programme and Characteristics o(Materials in Part II 160
A = wavelength of radiation
Diffraction will occur for the many different sets of atomic planes which can
occur in the crystal. A monochromatic X-ray beam will be diffracted at different
angles by the crystal structure of different minerals.
160
Chapter 6 Experimental Programme and Characteristics ofMaterials in Part II 161
The powder pattern X-ray data manual on cement minerals are according
with the Joint Committee on Powder Diffraction Standards ( JCPDS ) and the
Powder Diffraction File ( PDF ), and these were used to identify compounds, by
comparing known patterns from the manual with the test-result pattern until a
suitable match was obtained.
161
Chapter 7 Test Results and Discussion o[Part 11 162
CHAPTER SEVEN
7-1. INTRODUCTION
The goal of this chapter is to investigate the role of fly ash in HVF A concrete
III microstructure mechanisms due to pozzolanic reaction. Control concrete
specimens containing AS TM Type I cement with mineral admixture were also
investigated for comparison purposes.
In the present investigation, samples were taken from the same concrete test
specimens, and were tested at corresponding ages to correlate microstructure
development with strength increase. Details of concrete mix proportions and testing
procedures have been described in Chapter three and Chapter six respectively.
162
Chapter 7 Test Results and Discussion of Part 11 163
Pore size distribution is a very important parameter that will indicate the change in
pore size under different conditions, since it will be affected by experimental
parameters such as degree of hydration, curing temperature, ageing, water /
cementitious ratios, and cement composition [70,71,115]. Pore size distribution in
combination with strength may give a more clear idea what mechanism controls the
strength when the experimental parameters are changed.
In this section a detailed analysis of the pore size distribution curves will be
presented, and on its basis, some conclusions will be drawn about the relationship of
both types of curing conditions on the pore structure of HVF A concretes. Some
relations between pore structure and mechanical properties will also be presented.
Under both types curing conditions with different curing periods, the total
accessible pore volume and porosity of OPC and HVF A concrete are also shown in
Tables 7-1, 7-2, 7-3, and 7-4.
163
Chapter 7 Test Results and Discussion of Part II 16-1
0.081-----1------------r-------~
28days
,----
OJ) 0.06
~ 350C
S
j 350F
------;t(-
~ 0.04
350SF
.~ ~
1 '. ..............
~
350S
~ 0.02 ..............
".
'.
I O.~~
..•·· .. ··-·~·
..·-.... ·-~
..·...... ··~---·
.... ·~
.... --.. ~~~~
0.08r------r------------.....,--------
90days
~0.06 350C
e
j 350F
~ 0.04 350SF
j ~
350S
~ 0.02
C)
I O~~~~~-~~~
0.0025 0.05
.. ·..
10 300
Pore Diameter, micro-meter
0.08 ,---------r--------------,--------~
0.07
'"" 540days
~0.06
e 350F
'-' 0.05 ----¥--
j:!Z 0.04 350SF
~
~ 350S
.~ 0.03
]
80.02
0.01
0.05 10 300
Pore Diameter, micl1>-meter
3
Figure 7-1 Cumulative pore volume curves for the 350 kg/m binder content
OPC and HVFA concrete ( Water curing)
164
Chapter 7 Test Results and Discu.<,sion of Part II 165
28days
OIl 0.06
~ ! 350C
S
! 350F
• ---7(--
.................
.........
....
I~~fl350S
~
..........
.................
10 300
Pore Diameter, micro-meter
0.08 I------I""---------------r-------~
90days
~0.06 350C
S
j 350F
---7(--
~ 0.04
350SF
.~...................•......
~
.~ -+-
-3 350S
~ 0.02 ...................
0.05 10 300
Pore Diameter, micro-meter
0.08 r - - - - - - - - - . - - - - - - - - - - - - - - - r - - - - - - - - - - - - .
0.07
""' 540days
SO.06
S 350F
'-' 0.05 ~
j 0.04
~
350SF
------4-
~ 350S
'.g 0.03
'3
30.02
om
OL-~~~~~~~~~~~~~~~==~------~~--~
0.0025 0.05 10 300
Pore Diameter, micro-meter
3
Figure 7-2 Cumulative pore volume curves for the 350 kg/m binder content
OPC and HVFA concrete (Air curing)
165
Chapter 7 Test Results and Discussion o[Part 11 166
0.08 1 - - - - - -1 - - - - - - - - - - - - , - - - - - - - - - - -
28days
0.06
F
bO
::J
e !--~
~
450F
]
~ 0.04 I-t-
<I) i 450SF
I _____
.~ i
.5
='
I 4'~S
~ 0.02
0.08 r - - - - - - - - - r - - - - - - - - - - - - - , - - - - - - - - - - - - ,
90days
~0.06
450C
S
~
450F
] -t-
~ 0.04
450SF
.~
.5 4508
='
~ 0.02 --*---
0.08 r - - - - - - - - - , - - - - - - - - - - - - - - - , . . - - - - - - - - - - - - ,
0.07
,-.,
540days
~0.06
e 450F
'--' 0.05 ---t-
j 450SF
~ 0.04
~
----
450S
'i 0.03
'3
~ 0.02
om
0.05 10 300
Pore Diameter, micro-meter
3
Figure 7-3 Cumulative pore volume curves for the 450 kg/m binder content
OPC and HVFA concrete (Water curing)
166
Chapter 7 Test Results and Discussion o[Part JJ 167
0.08
.,
28days
OIl 0.06
~ 450C
e
0
450F
] --t---
~ 0.04
d)
;;.
'..\
\
I 450SF
j
;:) \; 450S
8 0 .02
\~ -----*"-
0.08 i - - - - - - - r - - - - - - - - - - - - - , - - - - - - - -
90days
~0.06 450C
e
.u 450F
] --t---
~ 0.04 450SF
d)
;;.
j
;:) 450S
8 0 .02
-~~
0.08 . - - - - - - - - - - - , - - - - - - - - - - - - - - - . . - - - - - - - - - - - - ,
0.07
,.-., 540days
5 0 .06
e 450F
'--' 0.05 -t--
0"
§ 4508F
~ 0.04
o 4508
;>
.~ 0.03
~ 0.02
0.01
0.05 10 300
Pore Diameter, micro-meter
3
Figure 7-4 Cumulative pore volume curves for the 450 kg/m binder content
OPC and HVFA concrete ( Air curing)
167
Chapter 7 Test Results and Discussion of Part 11 168
Table 7-1 Total accessible pore volume ofOPC and HVFA concrete-Water
curing
(Unit: mL/g)
Days 28 90 540
Table 7-2 Total accessible pore volume ofOPC and HVFA concrete-Air curing
(Unit: mL/g)
Days 28 90 540
168
Chapter 7 Test Results and Discussion o(Part 11 169
Days 28 90 540
Days 28 90 540
169
Chapter 7 Test Results and Discussion of Part 11 170
From the results, total cumulative pore size volumes of all concretes tend, as
expected, to decrease with increased hydration. At high degrees of hydration the
curves tend to become concave in the case of all concretes under water curing
condition. Meanwhile, the total accessible pore volume and porosity of OPC and
HVFA concrete toward to the lower value were found.
Under water curing, the concrete made with 350SF appeared to have more or
less the same cumulative pore volume of 350S, but the pore size distribution of this
porosity is quite different ; the former had most of its porosity in the coarse size
range, whereas that concrete of350S is mostly in the fine size range. These variation
of fine and coarse pores due to pozzolanic activity will be discussed in the next
section.
At the end of 90 days, for all concretes under water cunng, the total
cumulative pore volumes are lower than 28 days. For the long term pore size
distribution test at 540 days of all HVFA concretes under water curing, as expected,
the total cumulative volumes of the geometrically continuous pore were low than at
the ages of 28 days and 90 days. Meanwhile, the concrete made of 450SF had lowest
percentage in porosity when compared to all of the HVFA concretes. This indicates
that the amount of hydration products in 450SF were proportionately more than that
other HVFA concretes. The presence of the silica fume in the HVF A concrete was
thus instrumental in causing the beneficial effect observed on concrete strengths. This
also explained the reason why the crushing pattern of the specimen containing fly ash
or with both fly ash and silica fume was more ductile than that concrete without
above mineral admixture.
170
Chapter 7 Test Results and Discussion ofPart II 171
The results also show that the use of air curing regime was a remarkable
factor in determining the pore size distribution. These tends reflect a marked change
in the total cumulative pore volume ; meanwhile, the percentage of porosity are
increasing with continued air drying. Both the normal ope and HVF A concretes
showed higher accessible pore volumes compared to the corresponding water curing
concretes. Irrespective of normal ope concrete, at the age of 28 days, the concrete
made of350F had the total cumulative pore volumes lower than the other two HVFA
concretes, Mix-350SF and Mix-350S. The concrete made of mixes containing fly ash
as a partial replacement for both sand and cement, Mix-350S, Mix-450S had good
ability to resist the loss of the potential strength in the long term by air dried
condition.
From the results, at the end of 540 days the less the porosity of HVF A
concrete in water curing, the more is the value in the air regime. This is in agreement
with the preceding discussion in swelling and drying shrinkage of HVF A concretes.
The concretes made of 350SF, 450SF showed significant reduction in total pore
volume and, thus, had a good ability to resist swelling, but higher drying shrinkage
strain were also found on both types of concrete.
The maximum dv / d ( log d ) and the threshold diameter of ope and HVFA
concrete under both types curing conditions are shown in Tables 7-5 and 7-6. The
corresponding maximum dv / d ( log d ) against the threshold diameter on pore size
distribution are illustrated in the Figs 7-5, 7-6, 7-7, and 7-8.
From the results, the differential pore volume plots clearly demonstrated the
effect of HVFA concretes on the pore size distribution. At those ages under water
curing, although the total cumulative pore volumes of the HVF A concrete were much
higher than the normal ope concrete, the differential pore volume showed a higher
peak with a wider spread for the fine pores than that ope concrete. At the age of 28
days concrete made of 450S had much more pores concentrated at pore diameter
smaller than 0.023 pm in comparison with the ope concrete under water curing.
171
Chapter 7 Test Results and Discussion ofPart II 172
Under similar conditions, irrespective of normal OPC concretes, the concrete made of
450S had cumulative pore volume lower than the other HVFA concretes, meanwhile,
had most of its porosity spreading in the fine size range. The peaks of concrete mixed
with cement alone replaced partial by fly ash, 350F and 450F, occur at a significantly
larger pore diameter range than for the comparable concretes made with silica fume.
The maximum pore diameters of the concrete 350F, 350SF and 350S were at 0.0406
pm, 0.0286 pm, and 0.0244 pm respectively. Concrete made of 450F, 450SF, and
450S, the maximum continuous pore diameter were at 0.0481 pm, 0.0327 pm, and
0.023 pm respectively. From the results, however, the "maximum continuous pore
diameter" is a useful parameter in comprehending the concrete permeability. If
permeability is directly related to the maximum continuous pore diameter, then the
concretes of350S and 450S would be the least permeable.
At the end of 90 days, again, the total cumulative pore volumes also were
found to be lower in the concrete of350S, and had much more than the other HVFA
concretes. Surprisingly, concrete Mix-450SF had a lower cumulative pore volume,
meanwhile, gave highest differential pore volume for the fine pores and lowest
differential pore volumes for the coarse pores when compared to the normal OPC
and all HVF A concretes. This indicates the presence of many pores in the fine pore
range and a fewer number of pores in the coarse pore range for the concrete of
450SF in comparison with the concretes of 450S and 450F. Whereas, the concrete
made of 450S had higher cumulative pore volumes than the other concretes, the
differential pore volume is toward the bigger pore size distribution when compared to
the concrete of 450SF. The maximum continuous diameter of the concrete Mix-
450SF and Mix-450S were at 0.0107 pm and 0.0196pm. These variations in the
volume of fine pores in the concrete, although difficult to relate directly with concrete
properties, are nevertheless reflected in the strengths of corresponding concretes. The
addition of silica fume did, however, shift the distribution of pores toward the small
sizes. For the long term test at 540 days, the dv / d ( log d ) curves for both types
cementitious content concretes toward to the finer pore size distribution were found.
Although the concrete made of 350F and 450F had lowest total cumulative pore
volume when compared with the other HVFA concretes, the total coarse pores
172
Chapter 7 Test Results and Discussion ofPart II 173
volume had rather high. Reduction in the volume of large pore was observed with the
progress of the pozzolanic reaction. High concrete strengths were generally
associated with lower volume of large pores in the concrete. The distribution of pores
moved toward to the finer diameter and this explained the phenomena of gradual
increase in strength of concrete containing fly ash and silica fume, even though it had
higher total pore volume than concrete without fly ash or silica fume.
From the Figs 7-6 and 7-8, the different pore plots clearly demonstrated the
effect of air drying on the pore size distribution. All the concretes showed an increase
in number of coarse pores accompanied by a decrease in number of fine pores.
Irrespective of normal ope concretes, the concrete made of 350F and 350SF had
most of it porosity in the coarse size range compared to the concrete of 350S at the
age of 28 days. The maximum continuous pore diameter of the concretes 350F,
350SF, and 350S were at 0.0853 pm, 0.0718 pm, and 0.0229 pm respectively.
Although the total cumulative pore volume of the concrete 450S was much higher
than the concretes made of 450SF and 450F, the differential pore volume had shown
two peaks with a wider spread at the fine pores range. The maximum continuous
pore diameter were at 0.0573 pm for Mix-450F, 0.0553 pm for Mix-450SF, and
0.0531 pm, 0.0219 pm for Mix-450S respectively. At the end of 90 days, the
concrete made of 350F had a highest total cumulative pore volume and the
differential pore volume and showed the maximum continuous pore diameter at
0.0991pm , which is coarse pore range. The concrete 450S still had a lower
cumulative pore volume than that of concretes 450SF and 450F, but, the differential
pore volume had a peak concentrated at fine range. It is worth pointing out that the
concrete made of 450SF had a higher cumulative pore volume as well as the
differential pore volume had a peak in the coarse pores higher than 0.0751 pm, which
located in the coarse pore range. For the long term test on the air curing of HVF A
concretes, the dv / d ( log d ) curves for both types cementitious material content
concretes were found toward to the coarse pore size distribution. The above results
were found to be consistent with the discussion in engineering properties. In general,
engineering properties of concrete is influence by the volume of all voids in concrete,
capillary pores, gel pores, and entrained air, if present. From the result, it could be
173
Chapter 7 Test Results and Discussion ofPart 11 174
inferred that, for hydration of concrete to continue, it is sufficient to prevent the loss
of moisture from the concrete.
Table 7-5 The max. dv/d ( log d ) and threshold diameter of OPC and HVFA
concrete-Water curing
Days 28 90 540
Mix Max. Threshold Max. Threshold Max. Threshold
dv/d-Iog d diameter dv/d-Iog d diameter dv/d-Iog d diameter
( mL/g-llffi ) ( Ilffi) (mL/g-j..I.ffi) ( Ilffi) ( mL/g-llffi ) ( Ilffi )
350C 0.0592 0.0129 0.0621 0.0141 --- ---
350F 0.0745 0.0406 0.0514 0.0223 0.0729 0.0081
350SF 0.0873 0.0286 0.0657 0.0206 *0.0706 0.0102
**0.0708 0.0073
350S 0.0913 0.0244 0.0914 0.0178 0.0866 0.0086
450C 0.0556 0.0162 0.0589 0.0129 --- ---
450F 0.0988 0.0481 0.0790 0.0200 0.0677 0.0069
450SF 0.0651 0.0327 0.0725 0.0107 0.0412 0.0094
450S 0.0881 0.0230 0.0898 0.0196 0.0702 0.0116
Table 7-6 The max. dv/d (log d) and threshold diameter ofOPC and HVFA
concrete-Air curing
Days 28 90 540
Mix Max. Threshold Max. Threshold Max. Threshold
dv/d-Iog d diameter dv/d-Iog d diameter dv/d-Iog d diameter
( mL/g-llffi ) ( Ilffi) ( mL/g-llffi ) ( Ilffi) (mL/g-llffi ) ( Ilffi)
350C 0.0594 0.0128 0.0383 0.0170 --- ---
350F 0.0596 0.0853 0.0669 0.0991 0.0536 0.0932
350SF 0.0522 0.0718 0.0440 0.0182 0.0603 0.1151
350S 0.0607 0.0229 *0.0501 0.0503 0.0621 0.0781
**0.0439 0.0156
450C 0.0352 0.0134 0.0363 0.0142 --- ---
450F 0.0582 0.0573 *0.0464 0.0751 0.0628 0.1160
**0.0500 0.0193
450SF 0.0506 0.0553 0.0451 0.0863 0.0679 0.1173
450S *0.0649 0.0531 0.0514 0.0169 0.0703 0.0650
**0.0567 0.0219
Remarks : * - Peak 1 and ** - Peak 2
174
Chapter 7 Test Results and Discussion of Part 11 175
0.1 i---------r------------------
2 0.08 28days
·sI
350C
0.06
350F
I!o
g ~
350SF
"e 004
co· ~
,g 350S
'-'
~
"e 0.02
0.1 r-------------~------------------------,-------------~
~ 0.08 90days
~
·s~ 0.06
350C
350F
i t ~
-*-
350SF
: i
!0.04 I : ~
,g
'-'
!\
I :
350S
~ 1
:
\:
"e 0.02 i 1
... . ".""." ::
~
I
..............
0.0025
. . .-.~.--. . ~
o~~~==~~d . -~~ 0.05 10 300
Pore Diameter, micro-meter
0.1 , - - - - - - - - - - - - - - , - - - - - - - - - - - - - - - - - - - - - - - - - - - , - - - - - - - - - - - - - - -
0.09
~ 0.08
540days
¥0.07 350F
~ 0.06 -*-
Jio 350SF
80.05 ------4-
"e 0.04 350S
~
'-' 0.03
"e
> 0.02
"0
om
175
Chapter 7 Test Results and Discussion of Part 11 176
0.1 ,-----1-----------,-------
·s~ 0.06
350C
350F
t!o
'8 -----*-
350SF
-0 004 ~
~.
350S
~
-0 0.02 \
o~~~~~··--~-----=-----=---=----=----=---=----=---=---~-----=---=--~----~---~-~~~~~~
, 0.0025 0.05 10 300
Pore Diameter, micro-meter
90days
.s 0.08
(I)
350C
~
-0 0.02
0.05 10 300
Pore Diameter, micro-meter
0.1 r--------------,-------------------------,--------------~
0.09
i 0.05
3508F
----4--
3508
-0 0.04
]>
'-' 0.03
-0
~
-0 0.02
0.01
o~~~--~~~~~~~~==~~~~====----~
0.05 10 300
0.0025
Pore Diameter, micro-meter
3
Figure 7-6 Differential pore volume curves for the 350 kg/m binder content
ope and HVFA concrete ( Air curing)
176
Chapter 7 Test Results and Discussion o(Part 11 177
0.1 I-----Tr-----------,-------~
,..::. I 450SF •
4~S
"0 0.04
tlll
.2
"0
I ~:
:
> 0.02
"0
0.1r-~----------_,--------------------------,---------------~
90days
~ 0.08
450C
~...
·s 0.06
()
450F
~
0() -+-
,..::.
450SF
~ 0.04
.2 450S
......,
~
"0
>
"0 0.02
0.1 ,--------------,--------------------------,----------------,
0.09
a 0.08 540days
o
~ 0.07
450F
~ 0.06 -+-
-in 4508F
8 0.05
om
0.05 10 300
Pore Diameter, micro-meter
3
Figure 7-7 Differential pore volume curves for the 450 kglm binder content
ope and HVFA concrete (Water curing)
177
Chapter 7 Test Results and DiscussIOn uf Part II 178
O.l,-----i----------------_---,
~ 0.08 28days
"'Q)
S 450C
g
·s 0.06
- - I
450F
--tl.o
()
() -t--
450SF
I
~
-.;)
1;)/)
004
•
--
.sa 450S
.....-
-.;)
;.
-.;) 0.02
0.1 r--------------r-------------------------,---------------
~ 0.08 90days
"'Q)
S 450C
§
.~ 0.06 450F
--
I;)/)
() -t--
()
~
450SF
-.;)I;)/) 0.04
.sa
-.;)
;.
-.;) 0.02
--
450S
0.1 .--------------,---------------------------,-----------------,
0.09
~ 0.08
540days
¥... 0.07
450F
~ 0.06 -+-
~ 450SF
8 0.05
0.05 10 300
Pore Diameter, micro-meter
3
Figure 7-8 Differential pore volume curves for the 450 kglm binder content
ope and HVFA concrete (Air curing)
178
Chapter 7 Test Results and Discussion of Part 11 179
In comparison with total binder content on all of concretes under both types of curing
conditions are shown from Fig 7-9 to Fig 7-20. The results show that the total binder
content does not seem to have much effect on the pore size distribution. Concretes
3
made with 450 kg/m mixes were slightly lower cumulative pore volumes than that
3
concretes 350 kg/m mixes under both types curing conditions. It is importan
pointing out that, although concrete made of 450S had higher cumulative pore
volumes, the pore size distribution had similar trend 350S under water curing at 90
days. The maximum continuous pore diameter of both types of concrete approach to
fine pore size range, which is 0.0178 Ilm for Mix-350S, 0.0196 Ilm for Mix-450S
respectively. Similar results were found in air curing; the maximum continuous pore
diameter of Mix-450S had a finer pore size distribution at second peak than that of
concrete of350S.
It is interesting in the differential pore size plots, the concrete of both types of
binder content under the same curing condition and age, had more or less a similar
tendency in differential pore size curve. Porosity is a major factor in controlling
strength and modulus of elasticity of concrete systems. As demonstrated in preceding
discussion on engineering properties, the pulse velocity is controlled primarily by the
dynamic modulus of elasticity of the concrete. Test concretes from all the mixtures
with both types cementitious material contents had shown similar behaviours of pulse
velocity under the same conditions. From the results, there is no doubt that the pulse
velocity of concrete is also influenced by the volume of all pores in concrete.
179
Chapter 7 Test Results and Discussion ofPart II 180
'§
'-" 0.06
~'-" 0.06
g
~
Q)
~
E
~
, 350F ;
~ 0.04
, *
450C ~ 0.04
i 450F ,
~
.~ 0.02
~ ~
-"'\. .~ 0.02
1 \
~ .. t::-~
"3
~
E
u 0?0025 0.05 10 300 u 0~0~2:--::5-~0..J..0-5~.........I~="'"'10=--3WOO
Diameter, micro-meter Diameter, micro-meter
.....
$ 0.1 2o 0.1 .------::j.--------,---~
~ E
e0.08
o
e0.08
·s o
·s
~~J
~0.06 ~0.06 350F
8
~O.04 \!
450C 8 *
450F
,
I
I
~0.04 +
t .i
\
Q[) Q[)
..9 0.02 I,'
\.~ ..9 0.02
~
,..,._.
0~025 0.05
-
"0 10 300 0.05 10 300
Diameter, micro-meter Diameter, micro-meter
.....
2 O.l,----,----------,-----, ~ 0.1 ..------~----...,------,
8 E
e
o
0.08 e0.08
o
·s ·s 350S
~0.06 350SF ~0.06 -B-
8 ~
450SF '8 450S
~0.04 ~ 0.04 -.
"0 "0
tl() tl()
~
"0
~
"0 0~025 0.05 10 300 "0 0~025 0.05 10 300
Diameter, micro-meter Diameter, micro-meter
180
Chapter 7 Test Results and Discussion of Part II 181
""'
bIl
0.08 ~ 0.08 r-----,-------,---~
~'-../ 0.06 ~'-' 0.06
1,;) I:)~
S
;:l
~ 0.04
~
... @J
450C
S
;:l
~ 0.04
~
350F
-r-
450F
+
.~
1
u
0.02
0?0025
\::\."'. -
0.05
.... .........•.
.~
"3
U
S
;:l
0.02
O?oO'-2~5-~0......0-5~-~~....l10-*--3-t.l
10 300 OO
Diameter, micro-meter Diameter, micro-meter
2 0.1 ~
... 0.1 , - - - - , - - - - - - - , - - - - - ,
S S
8o 0.08 8o 0.08
's~0.06 '8
~
~0.06 350F
8 450C 8 *
450F
""' 0.04
'"C)
""' 0.04
'"C)
+
00
.Q 0.02 (; 00
.9 0.02
.... . .
-
0.05 10 300 0.05 10 300
Diameter, micro-meter Diameter, micro-meter
]
'-' 0.06
~'-' 0.06
~
I:)~
350SF 350S
;:l
~
§ -&
~ 0.04
~
~
0.04 450SF
~
...
450S
I 0.1
i
S
0.1 , . - - - , - - - - - - - , - - - - - - ,
8o 0.08 8o 0.08
's 350SF
'8 350S
~0.06 00 0.06 -&
oo ~ ----8
~
'"C)
0.04
450SF
""' 0.04
'"C)
...
450S
bIl bIl
.9 0.02 .9 0.02
'"C)
::2
> 10 300
'"C) O?o025 0.05 10 300 .; O?o025 0.05
Diameter, micro-meter Diameter, micro-meter
181
Chapter 7 Test Results and Discussion ofPart 11
182
'; 0.08 , - - - , - - - - - . , - - - - - - .
~'"-' 0.06
~~ 450F
~ 0.04
*
450F
~ +
.~ 0.02
'"3
S
8 0~OL2~5~~0~.0~5~~~~.....110~~-I3JOO
Pore Diameter, micro-meter
~ 0.1 r - - - , - - - - - - - - r - - - - - - ,
S
2
Q
0.08
·s
On 0.06 350F
'0
Q
r:. 0.04
*
450F
"'0
+
01)
...9 0.02
0.05 10 300
Pore Diameter, micro-meter
,-... 0.08 , - - - - . , . - - - - - - - r - - - - - ,
~'--' 0.06
<If
350SF 350S
~
450SF ~::> 0.04 -&
450S
-.
~
~ 0.02
~
0.05 10 300 u 0~025 0.05 10 300
Pore Diameter, micro-meter Pore Diameter, micro-meter
~ ~
'"Cl 0~025 0.05 10 300 '"C 0~025 0.05 10 300
Pore Diameter, micro-meter Pore Diameter, micro-meter
182
Chapter 7 Test Results and Discussion of Part 11 183
~
'-' 0.06 '-' 0.06
S~
;::l
"00.04 ~ ~~
~ 0.04
350F
+
~
~
.~
~ ~
450F
+
0.02 .~ 0.02
~ ~
§ E
;::l
u O~~~~--------~~=-~
u 0?0025 0.05 10 300 0.0025 0.05 10 300
Diameter, micro-meter Diameter, micro-meter
2 0.1 r-------,;------..,-----,
....
~ 0.1 r---,-------,---~
S E
e0.08
o e0.08
o
·s ·s
i O 06
. ~0.06
8
350F
*
450F
~0.04 ~0.04 +
'"..9 0.02
0/) '"
0/)
..9 0.02
~
'" 0?0025 0.05 10 300 0.05 10 300
Diameter, micro-meter Diameter, micro-meter
....
2 0.1 ,.----.,.....-----......,.-----, ~ 0.1
S E
e0.08
o
e0.08
o
'seo 0.06 350SF
'seo 0.06 350S
....... ~
8 ~
450SF 8 450S
:
0/) '"
0/)
183
Chapter 7 Test Results and Discussion o(Part 11
184
~ ~ :K
'-' 0.06 '-' 0.06
Q)' Q)'
S S i 350F
::I ::I
"00.04 ~ 0.04
~ i 4;F
1;: 1;: +
.~ 0.02 .~ 0.02
"3 "3
§ §
u 0?0025 0.05 10 300 u 0 ?oOL2~5~~O...l.0-5--":!::::::::~,,,"*,J..l0-~4"3..J
OO
Diameter, micro-meter Diameter, micro-meter
....
2 O.l r - - - - - - , - - - - - - , - - - - - , Q) 0.1 , - - - , - - - - - - , - - - - - ,
~
Q)
S
8 0.08
()
8 0.08
·s
()
'8
~0.06 ~0.06 350F
8
~0.04
8
~0.04
*
450F
+
-0 -0
01) 01)
.Q 0.02 .Q 0.02
~ ::2
-0 0~025 0.05 10 300 ~ 0~OL2~5-~0....L.0-5~+-~~.J.I0-~~300
Diameter, micro-meter Diameter, micro-meter
'-' 0.06
~'-' 0.06
ef 350SF ~' 350S
::I ~ -&
~ 0.04
1;:
450SF "0 0.04
~
1;:
...
450S
.~ 0.02 .~ 0.02
"'3
§ ~
u 0~025
8 0 L~---.l_=::::::~~==--..J
0.05 10 300 0.0025 0.05 10 300
Diameter, micro-meter Diameter, micro-meter
.... ....
~ O.l,----.,--------.------, ~ 0.1
S S
8 0.08
()
8 0.08
()
's '8
~0.06
--800 0.06
350SF
~
8 450SF
~0.04 ~0.04
01) 01)
.Q 0.02 .Q 0.02
-0 ::2
> ~ 0~025
-0 0~025 0.05 10 300 0.05 10 300
Diameter, micro-meter Diameter, micro-meter
184
('hapler 7 Test Results and Discussion of Part 11 185
,-., 008 r - - - - , - - - - - r - - - -
tll)'
~'--' 0.06
~~ 350F
~ 0.04 *
450F
~ +
~ 0.02
.B 0.1 r---r------...,-----
SI
8u 0.08
·s
~0.06 350F
u
r: 0.04
*+
450F
"0
tll)
.Q 0.02
'-'
~
~ 0~025 0.05 10 300
Pore Diameter, micro-meter
~ 0.08 , - - - , - - - - - - r - - - - - , ~ 0.08 , - - - - - , - - - - - - - , - - -
~ ~
~ ~
'-' 0.06 '-' 0.06
~"
/j)"
350SF 350S
~::> 0.04
~ 0.04
~
~
450SF
~
..
-e-
450S
! 0.02 j 0.02
~ ~
u 0~025 0.05 10 300 u 0~025 0.05 10 300
Pore Diameter, micro-meter Pore Diameter, micro-meter
jI-<
0.1 ~
I-< 0 .1 ,-----r------,-----,
S
8 0.08 I
80.08
.§ .§
-inu 0.06 350SF do 0.06 350S
u
~0.04
'\j
tll)
~
450SF '8
r: 0.04
"0
tll)
..
-e-
450S
185
Chapter 7 Test Results and Discussion ofPart II 186
As discussed in the preceding section, the pore structure will be divided into three
parts following closely the division defined by IUPAC : macro-pores ( d > 500 A ),
meso-pores ( 25 A", 500 A ), and micro-pores ( d < 25 A ). The effect of different
curing conditions on the total macro-pores, and meso-pores distribution of both
types binder contents are presented in Tables 7-7 and 7-8.
The results show that all the concrete mixtures with continuous water curing
exhibited a higher percentage meso-pores than the air drying concretes. Meanwhile,
the use of air curing of all concretes a slight reduction in meso-pore, and macro-pore
are just relatively increase with continued air drying.
Under water curing, however, the HVFA concretes had much greater meso-
pores, when compared to the normal OPC concrete at the ages of 28 days and 90
days under the same conditions. The trends in the macro-pores are just the opposite.
The larger percentage of macro-pores of normal OPC concrete reflects its "coarser"
pore structure. It is reasonable to suppose that increase in meso-pores with increased
curing age is the result of growth of hydration products during hydration reaction. On
this table, it can be clearly seen that the concrete of 450SF the meso-pores range and
much less in the macro-pores range at the age of 90 days. Meanwhile, at the end of
540 days, the concrete made of 350S and 450S had a highest percentage of meso-
pores compared to the other HVFA concretes.
Macro-pore into meso-pore due to reduction of their pore size by the growth
of hydration products into the available spaces between the unhydrated grains, which
were initially filled with water [71].
Data from the results also showed that an increase in the accessible pore
volume in the presence of concrete with air curing is associated with a decrease in the
number of fine pores, and an increase in the number of coarse pores when compared
to the wet specimens at the same conditions. This shift in pore structure is exhibited
in the differential pore volume curves as were already shown in above Figs 7-2, 7-4,
7-6, and 7-8.
186
Chapter 7 Test Results and Discussion of Part II
Table 7-7 The macropore and mesopore on ope and HVFA concrete-Water
curing
( Unit: % )
Table 7-8 The macropore and mesopore on ope and HVFA concrete-Air
curing
(Unit: %)
187
Chapter 7 Test Results and Discussion of Part 11 188
The pore size distribution in the presence of air curing concrete mixtures
changed from a narrower to a wider distribution with a shift in meso-pore diameter
toward to a macro-pore diameter. However, the variation in macro-pore, meso-pore
indicated the air drying has a marked effect on the arrangement of these growing
hydration products.
The relationship between the compressive strength, fe, and porosity of concrete, X, is
by the power equation :
b
fe = a X --------------- ( 7-1 )
where fe is the compressive strength in MPa, X is the porosity in %, and a and bare
constants.
Although both mixes with 350 kg/m3 and 450 kglm3 total binder content
concretes had shown similar pore size distribution under the same conditions, the
curing regime was the marked factor in determining the result of compressive
strength, and the pore size distribution. In general, the water cured concretes showed
higher values of compressive strength than those exposed to drying concretes ; it is
suggested that separate equations be developed for each curing condition.
The relation between porosity and compressive strength for all concretes
under different curing conditions is shown in Figs 7-21 and 7-22. The relationship is
based on the data at 28 days, 90 days, and 18 months for all concretes under water
and drying curing conditions. Regression analysis of the data shown in the figures
produced the following relationships :
188
Chapter 7 Test Results and Discussion of Part 11 189
120
350C 350F 350SF 350S
110
o X o o
~
~
90
• + •
t53 80
~
00 70
Q)
• • o
.....> +
rIl
rIl
Q)
60
~0
50
u
40
30
20
4 6 8 10 12 14 16
Porosity, %
120
350C 350F 350SF 350S
110 o o
D X
~
~
90
• + •
t53 80
~
00 70
Q)
.....>
rIl
~
60
a
§ 50
u
40
30 x + o
20
4 6 8 10 12 14 16
Porosity, 0/0
189
Chapter 7 Test Results and Discussion o(Part II 190
The correlation factor with the two equations is low. These data indicate a
significant correlation between porosity and compressive strength in concretes at the
95 % confidence level. A single equation describing the two parameters can be used
as given below, and is shown in Fig 7-23.
Again, the correlation factor between both parameters of equation is still low
and the equation relating the two is approximately the same for the water and air
drying curing concretes. All concretes have this same inverse dependency between
strength and porosity. Many workers [91,116] have made attempts to correlate
various mechanical properties with parameters related to pore structure, especially
total porosity. In fact, the effect of the degree of hydration on macro-pore and meso-
pore diameter obtained by Mercury Intrusion Porosimetry is shown in Tables 7-7 and
7-8. In most cases, very large changes occur in the diameter of the macro-pores and
meso-pores. The variation in macro-pores and meso-pores indicate that curing
condition and the degree of hydration have a marked effect on the arrangement of
these growing hydration products [70,71]. Generally, at a given porosity, smaller
pores lead a higher strength of the concrete. Regression analysis in Fig 7-24 shows
that the coefficient of determination ( r = 0.90 ), using meso-pore data in the
relationship, is very high. The results infer that high strength / performance concretes
are seen to be insensitive to the structure of the smaller pores, and are not to be
associated with the total porosity. As can seen in the preceding Tables 7-3, 7-4,
porosity over the whole size rang of pores has influence on concrete properties, yet it
is difficult to get an exact assessment of pore size distribution because Mercury
Intrusion Porosimetry measurement encompasses the whole size range and because it
is difficult to interpret experimental data. Thus, comparisons of porosity should be
made with care.
190
Chapter 7 Test Results and Discussion of Part 11 191
120
350C 350F 350SF 350S
no 0 X 0 0
~
d
90
• + •
o.
tas 80
.p
••
rn 70 0
•
.-Q)
;..
til
til
Q)
60
0
0 0
0
,• •
0
'P-"' o
§ 50 ( MOIST + DRY)
U
40
0
30 x +-t 0
20
4 6 8 10 12 14 16
Porosity, %
Figure 7-23 Relationship between compressive strength and porosity of ope
and HVFA concrete
80
350F 350SF 350S 450F 450SF 450S
X 0 0 + •
d 70
~ •
tas
~ 60
-
Q)
;..
.til
til
Q)
0
•
~ Y = 25.996 X
" 0.20
(r = 0.90) (MOIST + DRY)
8 50
+
40 L-~~~~~ __L-~-L~~__~~~~~~--~
10 20 30 40 50 60 70 80 90 100
Meso-pore, %
Figure 7-24 Relationship between compressive strength and meso-pore of
HVFA concrete
191
Chapter 7 Test Results and Discussion of Part II 192
The permeability test is based on "The Leeds Permeability Cell" [112], which
measures the coefficient of permeability by measurement of through flow of oxygen.
Permeability tests were performed on 50 mm diameter 30 ~ 40 mm cores samples
drilled from 100 x 100 x 500 mm prisms. The test specimens were subjected to
different curing and tested at the age of 28 days, 90 days, and 18 months. For each
curing three cored samples were used, the results are thus the average of three
readings. Mix proportions and test procedure described in section 6-4-2. Variations
in different replacement methods and total cementitious content also are investigated.
The relationship between the compressive strength and permeability are also
reported. A relation between the air permeability and porosity for HVF A concretes
was found.
Under water curing with various curing periods, the data on the air permeability of
HVFA concretes with both types of binder contents are given in Table 7-9 and
illustrated in Figs 7-25 and 7-26.
192
Chapter 7 Test Results and Discussion of Part II 193
Data from the results indicated that the coefficient of permeability was low in
water cured specimens at early ages for all HVFA concretes. Between the ages of 28
days and six months, obviously, the presence of silica fume in mixes 3508F and
4508F helped mixes 350F and 450F to develop a denser internal structure and a
lower permeability. Mixes 3508 and 4508 with large amounts of fly ash developed a
much denser structure and low permeability from early ages. As already stated in
section 5-5, the permeability of concrete is not only a simple function of its porosity,
but depends also on the size, distribution, shape tortuosity, and continuity of the
pores. The present investigation, from the preceding results of the pore size
distribution in Table 7-7, suggested that the permeability is strongly controlled by an
interconnecting network of capillary pores. At 28 days and 90 days under water
curing, the concrete made with 4508 had much lower macro-pores, compared to the
same condition of concrete 450F. The reduction in amount of macro-pores with
duration of specimens under water is correspondingly reflected in the permeability of
concrete.
193
Chapter 7 Test Results and Discussion orpar! II j;-/.
0.35 r--------------------~
Ne
'-"
-......
'C)
b
0.3
\
....r 0.25
5
......
(,)
~
8
u 0.2
C
.....
......
~~ 0.l5 \\~
~ :~---~
28 90 180 540
Age (Days)
0.3 , - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
450F 450SF 450S
• •
Ne 0.25
-sn 0.2
.....o
(,)
~
o
o
u 0.15
C
.....
......
.....
'io
§
o
O.l
~
28 90 180 540
Age (Days)
194
Chapter 7 Test Results and Discussion o[Part II 195
It is clear from those results that the permeability of the HVF A concrete was
directly related to the pozzolanic reaction of fly ash and silica fume in concrete,
which produces calcium silicate hydrate, tends to fill these unoccupied spaces to form
a product with decreased permeability and which also results in improved strength
[28]. These results were also found by Manmohan and Methta [37]. They showed
that additions of pozzolanic and cementitious admixtures such as fly ash to portland
cement were instrumental in causing pore refmement or transformation of large pores
into fine pores - a process which had a far reaching influence on the permeability of
the hardened cement concretes. Meanwhile, the use of fly ash to replace both cement
and sand partially might have a good combination effect on permeability.
In general terms, it is possible to say that the higher the strength of the
hardened cement concrete the lower its permeability - a state of affairs to be expected
because strength is a function of the relative volume of gel in the space available to it
[69]. Nevertheless, from the figures, the improvement in strength with ages of water
curing seems not to be correspondingly reflected in the permeability of concrete. For
the long term under water curing, the concretes exhibited slightly increasing
coefficient of permeability at the age of 18 months when compared to the
permeability with the same conditions at the age of six months. In comparing both
porosity ( Tables 7-7 and 7-8 ) and coefficient of permeability of concrete at the age
of 504 days, the concrete made of 450F has a higher macro-pore volume but a lower
coefficient of permeability. Whereas, for concrete made of 450S a lower percentage
in macro-pores was found. The corresponding coefficient of permeability for the
concrete made of 450S were high. It would be expected, as a first approximation,
that a sample with a higher porosity also would have a large permeability. However,
this generality does not take into consideration the possibility of and potential effects
of differences in pore size distribution. Research data [9] indicate the sensitivity of
high strength fly ash concrete to curing, especially for the most mature concrete. In a
mature concrete, the permeability depends on size, shape, and concentration of the
gel particles and on whether or not the capillaries have become discontinuous.
195
Chapter 7 Test Results and Discussion of Part 11 196
Days 28 90 540
196
Chapter 7 Test Results and Discussion ofPart 11 197
It is clear from these data that the permeability of the HVF A concrete was
directly related to the moisture at any give time. All results also indicate that concrete
curing in water is superior to that air drying curing. This advantage has also been
observed for both strength development rate and dynamic modulus of elasticity of
concrete.
The results indicate that the use of air curing of all HVFA concretes exhibited
a marked increase of permeability with continued air drying, even though the
concrete had higher strength. In general, the coefficient of permeability of concrete
incorporating both fly ash and silica fume as partial replacement for cement
performed lower than that concrete with cement alone replaced partially by fly ash.
197
Chapter 7 Test Results and Discussion of Part 11 198
In Figs 7-27 and 7-28, the coefficients of permeability, Per, are plotted against the
corresponding compressive strength, fe, of the data at the ages of 28 days, 90 days,
180 days, and 18 months for moist and dry cured specimens of both types of
cementitious material content HVF A concretes. Regression analysis of the data
shown in these figures produced the following relationships :
where fe is the compressive strength in MPa, and Per is the coefficient of permeability
in 10- 16 ( m2 ).
198
Chapter 7 Test Results and Discussion o(Part II 199
80r-----~~------------------------------
70
~
~
28
~ D I
water
" 60
ts:: 90
(\)
~ 50 0
.-
(\)
;;>
1Zl
1Zl
(\)
180
0
~40 540
0 B
U
30
o 0.1 0.2
16
0.3 0.4
Permeability Coefficient, 16 (nl )
Figure 7-27 Relationship between permeability and compressive strength of
HVFA concrete-water curing
80
28 540
EI
70
~
~
" 60
ts:: ~-
(\)
~ 50
.-(\)
;;>
1Zl
1Zl
E'~ 40 +++
0
u
30 +
20L---~--L---~--~--~--~--~---L--~--~
o 0.2 0.4 0.6
16 2
0.8 1
Permeability Coefficient, 16 (m )
199
Chapter 7 Test Results and Discussion of Part II 200
It is obvious from the Fig 7-27, the higher degrees of hydration of HVF A
concretes at the age of 540 days, the greater is the coefficient of permeability. In
view of the positive correlation factor ( r = 0.21 ) found in the equation relating to
the concrete at 18 months with water curing, it is suggested that the correlation
coefficient was very low. It is relieved that the compressive strength is still controlled
by the porosity, the more the meso-pores of the concrete, the stronger is the
compressive strength.
16 2
where Per is the coefficient of permeability of HVF A concrete in 10- m , Meso is the
meso-pore volume ofHVF A concrete in %, and a and b are constants.
200
Chapter 7 Test Results and Discussion of Part II ]01
the air cured specimens. Although these correlation coefficients are somewhat low,
there is a significant correlation at the 99 % confidence level between the coefficient
of permeability and the meso-pores of HVFA concretes with both types of
cementitious material contents under both curing conditions. Now, satisfactory
explanation is apparent for this, however it is likely to be connected to the
permeability of the hardened cement concrete when an adequate moist, and the
degree of hydration, and hence lower coefficient of permeability was obtained.
1.6 ,-----------------------~
o W ~ W W 100
Meso-pore, %
Figure 7-29 Relationship between permeability and meso-pores ofHVFA
concrete
201
Chapter 7 Test Results and Discussion of Part 11 202
[121]. Other important factors, such as the curing history, water / cement ratio,
aggregate characteristics, air content, cement type and fineness, and surface
carbonation may also an effect on absorption.
Under water curing for various periods, the results of the water absorption of HVF A
concretes with both types of cementitious material contents are shown in Table 7-11
and plotted in Figs 7-30 and 7-3l. When compared at equal 28 day and 90 day
strength for the wet-cured specimens the concretes made with 350SF, 450SF had
greater water absorption, whereas for the Mix-350S and Mix-450S, with fly ash both
as cement and sand replacement, had the lower water absorption. At the age of 180
days, the water absorption of all the HVF A concrete was reduced so that a lower
valued for both types of cementitious contents was measured. A more noticeable
reduction for concrete 350SF and 450SF, relative to the corresponding 350F and
450F, was observed.
202
Chapter 7 Test Results and Discussion of Part 11 203
Table 7-11 High volume fly ash concrete water absorption ( Unit: % )
For the long term wet curing, the water absorption of all HVF A concretes
have appeared similar behaviour to that air permeability. The concretes exhibited
slight increments in water absorption at the age of 18 months when compared to the
age of six months with the same conditions. However, the higher potential strength
and the consequent pozzolanic reaction of the concretes do exhibit good ability
characteristics to resist water penetration.
In moist concretes, pores of different size are always occupied so that the
absorption characteristics of HVF A concrete under water curing as well as drying
condition varied depending on its pore volume and pore size distribution. From the
preceding discussion on porosity, the meso-pore volumes of the concrete made of
350S are more than those of the concrete of 350SF and 350F. The result may
reasonably suggest therefore that water does not easily move through the very small
gel pores and that absorption is controlled by an interconnecting network of capillary
pores [13]. Furthermore, as mentioned previous section, using fly ash to replace both
cement and sand partially might have a combined effect of both separate
replacements.
203
Chapter 7 Test Results and Discussion o{Part II 204
3 r-----------------------------------------
350F 350SF 3508
/ 0 0
2.5
0.5L-J.----~--.L--'---'----'-....L.----------L---------'----L---'-.l.......J
2S 90 ISO 540
Age (Days)
3 ~--------------------------------------------
';:!(
0
§~
..... 2
e-o
rIl
~.... -+-==--~-
-----========= ~,--
1.5
2
~
1 •
0.5 L2...i.S----L-----'------'---'--9LO----l-iS-0------'----'-----'-5..i--J
40
Age (Days)
204
Chapter 7 Test Results and Discussion of Part 11 205
Another important consequence of the concretes' 450 kg/m3 mixes, at the age
of 18 months, was that the difference in water absorption was very small and
consistent trends could be identified.
From the results, the curing regime had a much more significant effect on the
water absorption than did the cementitious material, the concretes cured under dry
conditions having a considerably greater water absorption [122]. These increases in
the water absorption could be due to the change in microhardness and retrogression
of the concretes due to air curing regime. For the most part, concrete after mixing
205
Chapter 7 Test Results and Discussion of Part 11 206
with water becomes a solid / fluid suspension and flowability gradually disappears as
free water in concrete is gradually reduced. At the same time, the rigidity of the
concrete' becomes stronger and begins to develop its strength. The reduction of free
water is to caused by the formation of hydrates such as ettringite and C-S-H which
need a large amount of water. The inner structure, due to the formation of hydrates,
becomes a network which contributes to strength [13]. However, the pre-
conditioning regime by air curing may cause a variety of change on internal parts of
concrete. Changes in mechanical properties and porosity with time have been
discussed previously for these system. The present investigation has suggested that
this change in water absorption may due to an accumulation of dense hydration
products consisting of C-S-H, ettringite, Ca(OH)2, and hydrated aluminates at earlier
periods around the unhydrated cementitious grains during air curing [123]. It is
possible that the interaction of hydration products with fly ash or silica fume blends
partially would cause increased loss of moisture and would be expected to lead to
more extensive drying shrinkage cracking, which would be extended to form
continuous channels around the unhydrated cementitious grains with time, thus
allowing gel to imbibe more water. The increase of water absorption also helps to
explain the lower strength, dynamic modulus of elasticity, and pulse velocity in the air
cured concretes due to the combined effect of a higher permeability and water
absorption.
The relationship between absorption and the coefficient of permeability for both
curing conditions and cementitious material contents, at the age of 504 days is shown
in Fig 7-32. Regression analysis to establish the relationship between water
absorption and the permeability coefficient from the data shown produced the
following equation :
206
Chapter 7 Test Results and Discussion of Part II 207
2.5 r - - - - - - - - - - - - - - - - - - - - -_ _~
• 1'.
1
540 540
water air
0.5 oLL-'----L--'----L--------1---'-----'-------'-------'1
0.2 0.4 0.6 0.8
16 2
Penneability Coefficient, 10- (m )
207
Chapter 7 Test Results and Discussion orPart II 208
The following section describes the X-ray diffraction to identify the crystalline
phase and also to investigate the crystal structure of some of the phases. Routine
qualitative scans of the various mixtures readily identified the predominant phases,
calcium hydroxide, Ca(OH)2, and calcium silicate hydrate ( C-S-H gel ), together
with evidence of unhydrated concrete minerals such as silicon dioxide ( a-quartz ),
The test specimens were subjected to different curing conditions and analysed
at the age of28 days, 90 days, and 18 months. For comparison the hydrated powders
of normal ope and HVFA concretes are also included. These studies were done for
the OPC and HVF A concretes made of 450 kg/m3 mixes. All the samples were mixed
and prepared as has been described in Chapter three and Chapter six
The hydration products of normal OPC and HVF A concretes with and without silica
fume under water curing with curing time are seen from the X-ray diffraction patterns
in Figs 7-33, 7-34, 7-35 and Tables 7-13, 7-14, 7-15.
In the above figures all the X-ray diffractograms have been plotted with
identical scales for both intensity ( I ) and 28, where 8 is the Bragg's angle. Overall
from the figures, a maximum intensity ( 100 % ) peak corresponding to the a-quartz
appears at 3.351 A for both OPC and HVFA concretes throughout the curing
hydration period.
208
Chapter 7 Test Results and Discussion of Part 1I 209
and perhaps carbo aluminate phase also appear in all samples independent of the
different replacement methods. The a-quartz is believed to be from the aggregate
used for compounding the concrete while, the calcite and carboaluminate phase might
be as a result of the carbonation of portlandite during curing.
However, for all mixtures, the intensity of the X-ray reflections corresponding
to the unhydrated compounds strongly decreased as a result of the hydration reaction
under water curing regime.
209
Chapter 7 Test Results and Discussion o(Part II 210
210
Chapter 7 Test Results and Discussion of Part II 211
211
Chapter 7 Test Results and Discussion of Part II 212
212
Chapter 7 Test Results and Discussion of Part II 213
213
Chapter 7 Test Results and Discussion ofPart 11 214
o
•
2000~
~
I
1I
1
I
450C
1000
450F
1000 o, N
x
U
I
o
xU
UJ
::r
U l
d
N
-g ::r
;::l
'j'"
U
d
o, o
/1 "
o ~-LJLJ t i , I I I I I : I . i
2000
II
450SF
N
1000 ::t:
U ~'j'"U
::r
d
0,
Ii
1000
450S o,
o o
" "
o, o
II
50 60 70
10 20 30 40
Degrees 2-Theta
Figure 7-33 X-ray ditTractograms for ope and HVFA concrete-Water curing
( 28 days hydration)
214
~IV;
IV .."
C>
C> C>
C>
C>
'"
o
o
C>
0
IV
C>
C)
C>
C>
o
C>
C> o C> C> C> C>
o 0 0.0 C> C>
C) I I I '. ------.L_~J _ _ L~__'__ ~
](-L- l--------.l~ ,~ I __ ~~_..l- __ --1...- _ _ _ _ L .. \ ---'--
_.
~
(JQ
.t>-
V>
o
.t>-
V>
o
.t>-
V>
0
.t>-
V>
o
91:::1
C/l ~
=
'"'t
~
C/l "T1 "T1 ()
""l
'-l
-..l
I
<> CH <> CH -j ==-=- <> CH ~
~ IV
~
~ C)
:::tl
_<> (x-quartz u-quartz
[~~
<> a-quartz u-quartz
>< ~
ll~
I :::!
'"'t
~
~ ~
::\
-.
Q.
a-quartz (x-quart/. a-quartz u-quanz =c-"-co~~'======~---
s:;.,.
tJ
~ .oO .0 ~ ~C> C;;'
=f"')
..... ~<>
()
:::!
V,
-e
(..>
Q o V,
(JQ C3
::\
\0 ;i a
-,
0 CH CH
~
~o
~
.0
CH
Q. ....,
=
'<
~
~
'"'t 0
(I)
-l~-
~<> a-quartz _0 u-quartz .C> a-quartz cO a-quartz
-.
~
N
...-
VI =-
'<
Q.
0
~
CO
iil
(I)
Ul~
~<> a-quartz .0 a-quanz _0 a-quanz a-quartz
(j ",0 .0
a-quartz
_. =
;i
.....
Q =
I
--l
::r
(I)
iii
~<>
a-quartz
a-quartz °
a-quartz
(x-quartz cO
a-quartz
a-quaru a-quaru
-~
Q.
=
== a-quartz
I f <>0° o.-qllanZ Cl-quartz
~~
0
u-quartz
CH -~Q CH
>
f"')
Q
=f"')
'"'t
.....
~
~
U'I
C)
'"
coO U-quartz
(x-quartz
r~ ~
L" 0
o
"-G""'
(x-quan/.
l
r
1! ~
C
0 '-"0
CH
u-quartz
"-G"""'
,
r'r .o - CH
«-qllart/
-ll~~"
"-quan/.
~
=
.....
~
'"'t Ol .::~> (x-qllarv ~r-=-::~ ~, «-qllart/
t ;;. "'---..
(x-qu;lrl/ l
--1~7- <> (L-quanl
to
f"') C)
I'
=
::!. ;0""
=
(JQ ~.::' u-quartz
"-G"'fu ~~j ~
</
(l-quartz
I~
o "C .... <=-~
;;. y' . «-quartz ~ ~';c: a-quartz -1 5.. '=''''''0 Cl-qlU1rtz
""~' a-qllart/
-oJ II
o
Chapter 7 Test Results and Discussion o{Part II 216
I)
)
3000-1
450F
1000
o,
450SF
1000j 450S
o, "
1:
1
l
10 20 30 40 50 70
Degrees 2-Theta
216
Chapter 7 Test Results and Discussion of Part 11 217
Under air curing, HVFA concretes with 450 kg/m3 mixes the X-ray diffraction
patterns tested at the end of 18 months are shown in the Fig 7-36 and Table 7-16.
The results show the remarkable influence of the air dried curing on hydration
products which are responsible for mechanical change.
Much of the carboaluminate phase and high intensity peak were recorded by
X-ray diffraction analysis on all HVF A concretes. CaC03 hydrates noticeably as is
evidenced by the persistent presence of the peak at 42.36 28, 42.38 28, and 42.4 28
in Mix-450F, Mix-450SF, and Mix-450S respectively.
From the comparative analysis of Fig 7-36 marked differences can be seen in
the degree of alteration caused by the carbonation. Although the combination of silica
fume and fly ash formed the silicic acid gels, carbonation results in the formation of
CaC0 3 through the intermediate formation of carboaluminate hydrates and by direct
reaction between calcium hydroxide, Ca(OH)2, and CO2 [103].
217
Chapter 7 Test Results and Discussion of Part II
218
4000 Start: 10° End:70° Step:.02° Speed:2°'min Time/Step:.01° Wavelength:1.541838A (Cu)
o,
3000-,
J
J
!
2000~
1000~ 450SF
iI o, o
o
u
u
Ll
Ll :r:
u ~
1'" :r: o
d U
'"
o I
o
Il
450S o
u
u'" N
400 t::
:r:
u
'" oIt
.
o
o,.
o
I
o, 0,
oI I I I I I I I I I 1 I r I 1 1 I I I r I I I I I I I
10 20 30 40 50 60 70
Oegrees 2-Theta
219
Chapter 7 Test Results and Discussion of Part 11 220
7-6. CONCLUSION
From the results of the microstructure analyses on HVF A concretes, the following
conclusions on the pore size distribution and hydrated products on OPC and HVF A
concrete, and the mechanisms controlling the permeability and water absorption are
presented.
1. The curing regime has shown a big influence on the microstructure development
ofHVFA concrete.
2. There exists an approximate indirect relationship between strength and pore size
of hardened concretes.
3. Reduction in the volume of large pores was observed with the progress of the
pozzolanic reaction. Higher concrete strengths were generally associated with
lower volume of large pores in the concrete.
4. The addition of silica fume to the mix decreased the accessible pore volumes for
HVF A concrete in comparison with the concrete without silica fume.
6. The higher potential strength of mature concrete are very sensitive to curing
temperature ( pre-oven dry, at 105 0 C, 24 hours) and degree of hydration.
7. The higher permeability and water absorption of air-cured HVF A concretes are
explained by a poor dispersion of the precipitated hydration products during the
hydration process.
220
Chapter 7 Test Results and Discussion o(Part 11 ]21
10. Decrease in the levels of calcium hydroxide phase were seen with advancing
water curing age in all the HVFA concretes, providing evidence for pozzolanic
reactivity of the fly ash. Furthermore, the calcium hydroxide phase is completely
consumed in HVF A at the age of 18 months.
11. Retrogression of HVF A concrete under air cunng IS associated with the
formation of CaC03 from Ca(OH)2.
221
Chapter 8 Conclusions and Recommendations {or Future Research
CHAPTER EIGHT
The conclusions drawn from this investigation are summarized at the end of each
part, but the overall major conclusions derived from this study are given below.
2. With HVF A concrete, increasing the total binder content brings little benefit in
terms of strength development, pulse velocity, and pore size distribution. The
concrete with the lower binder content had, however, a marginally higher pulse
velocity. Using a low binder content is not only economical per se, but also leads
to overcoming the undesirable effects of drying shrinkage.
4. Although the early strength of concrete made with high volume fly ash was low,
all HVF A concretes developed compressive strength at 28 days in the range 30 to
222
Chapter 8 Conclusions and Recommendations tor Future Research
40 MPa, and 40 to 65 MPa at 90 days. The use of fly ash to replace both cement
and sand simultaneously resulted in considerably higher strength values due to the
decreased w / b ratio, and reduced loss of early strength. The presence of silica
fume changes the strength of HVF A concrete due to the combined effect of
accelerating hydration reactions with microstructural changes.
t
High Strength ;
Low Porosity; Fly Ash High Early Strength ;
Low Swelling; Low Shrinkage;
Good Economics. Good Economics.
Partial
Fly ash Replacement
with of
Silica Fume Cement and Sand
223
Chapter 8 Conclusions and Recommendations (or Future Research 22-1
6. The ettringite formed in the early stages of hydration accelerated swelling. This
suggests that swelling begins at a lower critical degree of hydration. A
considerably lower swelling was observed in the HVF A concrete containing silica
fume, and this resulted in a concomitant decrease in the total accessible pore
volume and porosity.
7. Low swelling of HVF A concrete results in higher drying shrinkage. The latter
6
ranged from 550 x 10-6 to 640 X 10- mlm for all HVFA concretes at 504 days.
8. The dynamic modulus results were generally high, with values of the order of 55
GPa at 90 days for HVFA concrete with silica fume. Due to high pozzolanic
activity, the percentage increment in dynamic modulus from 28 days to one year
of HVFA concrete is about 3 ~ 4 times that of the normal OPC concrete. At the
end of 18 months, the results are higher than the ACI Building Code 318-89
calculated values by about 50 % for the 450 kglm3 mixes.
9. The pulse velocity of HVF A concrete reached stable values by about six months.
Both dynamic modulus and pulse velocity give good indication of the extent of
deterioration of concrete affected by prolonged air drying. The average
percentage loss of dynamic modulus or pulse velocity for all HVF A concretes
was very low, about 3 0/0, when compared to the same concretes air cured for 28
days.
10. Concretes stored in air showed higher carbonation depths than those stored in
water. Accompanying carbonation, HVFA concrete showed decomposition of
hydration products, and as a result, concrete becomes porous.
11. The results from the air permeability and water absorption tests are, to some
extent, related to the results from the pore size distribution test. Both these tests
thus seem suitable for predicting changes in the pore size of concrete systems.
The HVF A concrete with wet curing showed low water absorption and
16 16 2
permeability coefficient. The latter ranged from 0.117 x 10- to 0.305 x 10- m
16 16 2
for HVFA concrete cured for 28 days, and from 0.102 x 10- to 0.151 X 10- m
for HVF A concrete cured for 90 days.
Chapter 8 Conclusions and Recommendations {or Future Research 225
12. The results show that porosity is a major factor in controlling the engineering
behaviour of HVF A concrete. There exists an approximative indirect relationship
between strength and pore size of hardened concretes.
13. X-ray diffraction studies made it possible to gain a better understanding of the
microstructural changes induced by these procedures. The formation of
predominant ettringite / gypsum leads to swelling / drying shrinkage. It was
shown that an increase in curing leads to a change in the microstructure of C-S-H
hydrates. Decrease in the levels of calcium hydroxide ( CH ) was seen with
advancing curing age in all the HVFA concretes, providing evidence for
pozzolanic reactivity of the fly ashes.
During this study several interesting points have come up, which could not be
investigated because of the limitations of time. For further work in this direction, the
author makes the following suggestions.
4. Studies oflong-term creep under different curing conditions will be very useful.
5. The use offly ash to replace both cement and sand has improved early strength. It
is necessary to investigate the effect of different exposure regimes such as,
225
Chapter 8 Conclusions and Recommendations {or Future Research 226
sodium chloride solution and sulphate solution, and evaluate the Iong-tenn
durability characteristics ofHVF A concrete.
226
References ]]7
References
6. Swamy, R. N., "Fly Ash Concrete Potential without Misuse", Materials and
Structures, No. 23, 1990, pp. 397-411.
227
References 228
11. Kokubu, M., "Fly Ash and Fly Ash Cement", Proceeding of the Fifth
International Symposium of the Chemistry of Cement, Tokyo, 1968, The
Cement Association of Japan Part IV, pp. 75-102.
13. Mindess, S. and Young, J. F., "Concrete", Prentice-Hall, Inc. Englewood Cliffs,
New Jersey, 1981.
15. Jahren, P., "Use of Silica Fume in Concrete", First International Conference on
The Use of Fly Ash, Silica Fume, Slag and Other Mineral By - Products in
Concrete, July 31 - August 5 1983, pp. 625-642.
16. Lamond, J. F., "Twenty Five Years Experience Using Fly Ash in Concrete", First
International Conference on The Use of Fly Ash, Silica Fume, Slag and Other
Mineral By - Products in Concrete, Canada, July 3 1 - August 5 1983, pp. 47-69.
17. Guillaume, L., "Pozzolanic Activity of Fly Ash in Portland Cement and Slag
Cement", Silicate Inds., Vol. 28, No.6, 1963, pp. 297-300.
228
References 229
19. Kokubu, M. and Yamada, 1., "Fly Ash Cements", The VI International Congress
on The Chemistry of Cement, Moscow, September 1974, pp. 1-47.
20. Clifton,1. R., Brown, P. W. and Geoffery Frohusdorff, "Reactivity of Fly Ashes
with Cement", Cement Research Progress, 1977, pp. 321-341.
21. Swamy, R. N., Sami, A. R. Ali. and Theodorakopoulos, D.D., "Early Strength
Fly Ash Concrete for Structural Applications", ACI Journal, September -
October 1983, pp. 414-4213.
23. Olek, 1. and Diamond, S., "Proportioning of Constant Paste Composition Fly
Ash Concrete Mixes", ACI Materials Journal, March - April 1989, pp. 159-166.
24. Berry, E. E. and Malhotra, V. M., "Fly Ash for Use in Concrete - A Critical
Review", ACI Journal, March - April 1980, pp. 59-73.
25. Hopkins, C. 1. and Cabrera, 1. G., "The Influence of Pulverised Fuel Ash on the
Workability of Concrete", Second International Conference on Ash Technology
and Marketing, Barbican Centre, London, September 16th - 24th 1984, pp. 393-
397.
26. Sturrup, V. R., Hooton, R. R. and Clendenning, T. G., "Durability of Fly Ash
Concrete", First International Conference on The Use of Fly Ash, Silica Fume,
Slag and Other Mineral By - Products in Concrete, July 31 - August 5 1983, pp.
71-86.
27. Langley, W. S., Carette, G. G. and Malhotra, V. M., "Strength Development and
Temperature Rise in Large Concrete Blocks Containing High Volumes of Low -
Calcium (ASTM Class F) Fly Ash", ACI Materials Journal, July - August 1992,
pp. 362-368.
229
References 230
29. Cripwell,1. B., Brooks, 1. 1. and Wainwright, P. 1., "Time Dependent Properties
of Concrete Containing Pulverised Fuel Ash and a Superplasticizer", Second
International Conference on Ash Technology and Marketing, Barbican Centre,
London, September 16th - 21st 1984, pp. 313-320.
30. Thomas Michael, D. A and Matthews, 1. D., "Performance of Fly Ash Concrete
in UK Structures", ACI Material Journal, Vol. 90, No.6, Nov. - Dec. 1993, pp.
586-593.
32. Bamforth, P. B., "In Situ Measurement of the Effect of Partial Portland Cement
Replacement Using Either Fly Ash or Ground Granulated Blast Furnace Slag on
the Performance of Mass Concrete", Proc. lnstn. Civil Engineers, Vol. 69" Part
2, Sept. 1980, pp. 777-800.
35. Tikalsky, P. 1. and Carrasquillo, R. L., "Influence of Fly Ash on the Sulphate
Resistance of Concrete", ACI Materials Journal, Vol. 89, No.1, Jan. - Feb.
1992, pp. 69-75.
230
References 231
36. Rezansoff, T. and Stott, D., "Durability of Concrete Containing Chloride Based
Accelerating Admixtures", Canadian Journal of Civil Engineering, Vol. 17, 1990
pp. 102-111.
37. Manmohan, D. and Mehta, P. K., "Influence of Pozzolanic, Slag, and Chemical
Admixtures on Pore Size Distribution and Permeability of Hardened Cement
Paste", Cement, Concrete, and Aggregates, Vol. 3, No.1, 1981, pp. 63-67.
39. Tikalsky, P. 1. and Carrasquillo, R. L., "Fly Ash Evaluation and Selection for use
in Sulphate Resistance Concrete", ACI Materials Journal, Vol. 90, No.6, Nov. -
Dec. 1993, pp. 545-551.
43. Georges, C., Alain B., Chevrier, R. L. and Malhotra, V. M., "Mechanical
Properties of Concrete Incorporating High Volumes of Fly Ash from Sources in
the US", ACI Material Journal, Vol. 90, No.6, Nov. - Dec. 1993, pp. 535-544.
231
References 232
46. Buttler, F. G., Deeter, M. H. and Smith, G. R., "Studies on the Desiccation and
Carbonation of Systems Containing Portland Cement and Fly Ash" , First
International Conference on The Use of Fly Ash, Silica Fume, Slag and Other
Mineral By-Products in Concrete, Canada July 31 - August 5 1983, pp. 367-380.
50. Dhir, R. K., Jones, M. R. and Seneviratne, A. M. G., "Diffusion of Chloride into
Concrete, Influence of PF A Quality", Cement and Concrete Research, Vol. 21,
No.6, November 1991, pp. 1092-1102.
51. Haque, M. N., Kayyali, O. A. and Gopalan, M. K., "Fly Ash Reduces Harmful
Chloride Ions in Concrete", ACI Materials Journal, Vol. 89, No.3, May - June
1992, pp. 238-241.
232
References 233
54. Xiaofeng, C., Shanglong, G., David, D. and Steven, L. M., "Role of Silica Fume
in Compressive Strength of Cement Paste, Mortar, and Concrete", ACI
Materials Journal, Vol. 89, No.4, July - August 1992, pp. 375-387.
55. Gj<j>rv, O. E., "Durability of Concrete Containing Condensed Silica Fume", First
International Conference on The Use of Fly Ash, Silica Fume, Slag and Other
Mineral By- Products in Concrete, Canada, July 31-August 5 1983, pp. 695-709.
56. Sellevold, E. J. and Radjy, F. F., "Condensed Silica Fume in Concrete; Water
Demand and Strength Development", First International Conference on The Use
of Fly Ash, Silica Fume, Slag and Other Mineral By - Products in
Concrete,Canada, July 31 - August 5 1983, pp. 677-694.
58. Swamy, R. N., "Design for Durability and Strength Through the Use of Fly Ash
and Slag in Concrete", unplished, March 1997.
233
References 234
62. Aimin, Xu. and Sarkar, S. L., "Microstructual Development in High Volume Fly
Ash Cement System", Journal of Material in Civil Engineering, Vol. 6, NO.1,
February 1994, pp. 117-137.
66. Goldman, A. and Bentur, A., "Bond Effects in High Strength Silica Fume
Concretes", ACI Material Journal, V. 86, No.5, Sept. - Oct. 1989, pp. 440-447.
68. Kesler, C. E., "Control of Shrinkage, Creep and Temperature Volume Changes
Through Variations of Material Properties", State of Art Report, Technical
Committee No. 25 : Creep, Shrinkage, and Temperature Effects, ASCE-IABSE
Joint Committee on Planning and Design of Tall Building, 15 January 1972, pp.
1-11.
234
References 235
71. Bentur, A., "Effect of Curing Temperature on the Pore Structure of Tricalcium
Silicate Pastes", Journal of Colloid and Interface Science, Vol. 74, No.2, April
1980, pp. 549-560.
72. Haque, M. N., Langan, B. W. and Ward, M. A., "High Fly Ash Concretes", ACI
Journal January - February 1984, pp. 54-59.
73. Detwiler, R. J. and Mehta, P. K., "Chemical and Physical Effect of Silica Fume
on the Mechanical Behaviour of Concrete", ACI Material Journal, Vol. 86, No.
6, November - December 1989, pp. 609-614.
74. Baalbaki, W., Aitcin, P. C. and Ballivy, G., "On Predicting Modulus of
Elasticity in High Strength Concrete", ACI Material Journal, Vol. 89, No.5,
Sept.- Oct. 1992, pp. 517-520.
76. Hobbs, D. W., "The Dependence of the Bulk Modulus, Young's Modulus,
Creep, Shrinkage and Thermal Expansion of Concrete upon Aggregate Volume
Concentration", Materiaux Et Constructions, Vol. 4, No. 20, 1971, pp. 107-
114.
77. Lobo, C. and Cohen, M. D., "Effect of Silica Fume on Expansion Characteristics
of Expansive Cement Pastes", ACI Material Journal, Vol. 89, No.5, Sept.- Oct.
1992, pp. 481-490.
235
References 236
78. Popovics, S., Rose, J. L. and Popovics, J. S., "The Behaviour of Ultrasonic
Pulses in Concrete", Cement and Concrete Research, Vol. 20, 1990, pp. 259-
270.
79. Litvan, G. G. and Meyer, A., "Carbonation of Granulated Blast Furnace Slag
Cement Concrete During Twenty Years of Field Exposure", Fly Ash, Silica
Fume, Slag, and Natural Pozzolans in Concrete, SP-91, American Concrete
Institute, Detroit, 1986, pp. 1445-1462.
80. Kokubu, M. and Nagataki, S., "Carbonation of Concrete with Fly Ash and
Corrosion of Reinforcement in 20 Years Tests", Fly Ash, Silica Fume, Slag and
Natural Pozzolans in Concrete-Proceeding of the Third International
Conference, SP-1l4, Vol. 1, American Concrete Institute, Detroit, 1989, pp
315-329.
81. Ewerstson, C. and Prtersson, P. E., "The Influence of Curing Conditions on The
Permeability and Durability of Concrete Results from a Field Exposure Test",
Cement and Concrete Research, Vol. 23, 1993, pp. 683-692.
82. Malami, C., Batis, G., Noulooumbi, H. and Kaloidas, V., "Influence of
Pozzolanic and Hydraulic Cement Additions on Carbonation and Corrosion of
Reinforced Mortar Specimens", Proceeding of International Conference held at
the University of Sheffield, 24-28 July 1994, Edited by R. N. Swamy, Vol. 2 pp.
668-682.
83. Nagataki, S., Ohga, H. and Kim, E. M., "Effect of Curing Condition on the
Carbonation of Concrete with Fly Ash and the Corrosion of Reinforcement in
long-term Tests", Fly Ash, Silica Fume, Slag and Natural Pozzolans in Concrete,
SP-91, American Concrete Institute, Detroit, 1986, pp 1445-1462.
84. Afridi, M. U. K., Chaudhary, Z. U., Ohama, Y., Demura, K. and Iqbal, M. Z.,
"Carbonation Resistance of Powered and Aqueous Polymer-Modified Mortars
and Effects of Carbonation on Their Pore Structures", Proceedings of
236
References 237
85. Dalgleish, B. 1., Pratt, P. L. and Moss, R. 1., "Preparation Techniques and the
microscopical Examination of the Portland Cement Paste and C3 S", Cement and
Concrete Research, Vol. 10, No.5, 1980, pp. 665-676.
86. Dent Glasser, L. S., Lachowski, E. E., Mohan, K. and Taylor, H. F. W, "A
Multi-Method Study of C3 S hydration", Cement and Concrete Research, Vol. 8,
No.6, 1978, pp. 733-740.
88. Bentur, A., Berger, R. L., Lawrence, F. V., Milestone, N. B., Mindess, S. and
Young, 1. F., "Creep and Drying Shrinkage of Calcium Silicate Pastes III A
Hypothesis of Irreversible Strains", Cement and Concrete Research, Vol. 9, No.
1, 1979, pp. 83-96.
89. Roy, D. M., Gouda, G. R. and Bobrowsky, A., "Very High Strength Cement
Pastes Prepared by Hot - Pressing and Other High - pressure Techniques",
Cement Concrete Research, Vol. 2 No.3, 1972, pp. 349-366.
90. Rossler, M. and Odler, I., "Investigation on the Relationship between Porosity
Structure and Strength of Hydrated Portland Cement Pastes I. Effect of
Porosity", Cement Concrete Research, Vol. 15 No.2, 1985, pp 320-330.
91. Watson, K. L., "A Simple Relationship between the Compressive Strength and
Porosity of Hydrated Portland Cement", Cement Concrete Research, Vol. 11,
1981, pp. 473-476.
237
References 238
93. Sersale, R., Cioffi, R., Frigione, G. and Zenone, F., "Relationship between
Gypsum Content, Porosity, and Strength of Cement", Cement and Concrete
Research, Vol. 21, No.1, 1991, pp. 120-126.
94. Goto, S. and Roy, D. M., "The Effect of W / C Ratio and Curing Temperature
on the Permeability of Hardened Cement Paste", Cement and Concrete
Research, Vol. 11, 1981, pp. 575-579.
97. Darmawan, L., Richard, L. B. and Young, J. F., "Simple Method for Measuring
Water Permeability of Concrete", ACI Materials Journal, Vol. 86, No.5,
September - October 1989, pp. 433-439.
98. Lea, F. M., "The Chemistry of Cement and Concrete", 1956, Second Edition.
100.Diamon, M., Abo-el-enein, S. A., Hosaka, G., Goto, S. and Kondo, R., "Pore
Structure of Calcium Silicate Hydrate in Hydrated Tricalcium Silicate", Journal
of The American Ceramic Society, Vol. 60, Mar. - Apr. 1977, pp. 110-114.
102. Jenning, H. M. and Pratt, P. L., "On the Hydration of Portland Cement",
Proceeding of British Ceramic Society, 1979, pp. 179-193.
238
References 239
103.Hansen, T. C., Radjy, F. and Sellevold, E. 1., "Cement Paste and Concrete" ,
Building Materials Laboratory, The Technical University of Denmark, 1973, pp.
233-268.
1l0.Shi, D. and Winslow, D. N., "Contact angle and Damage During Mercury
Intrusion into Cement Paste", Cement and Concrete Research, Vol. 15, 1985,
pp. 645-654.
239
References 2-10
112. Cabrera, 1. G. and Lynsdale, C. 1., "A new Gas Permeameter for Measuring the
Permeability of Mortar and Concrete", Magazine of Concrete Research, Vol. 40,
No. 144, September 1988, pp. 177-182.
117 .Jose Anagel Lechuga Albala and Sebastla Alegre Rossello, "Concrete
Impermeability Study", Second International Conference on Ash Technology
and Marketing, Barbican Centre, London, September 1984, pp. 387-389.
1I9.Dhir, R. K. and Byars, E. A., "Pulverised Fuel Ash Concrete: Near Surface
Absorption Properties", Magazine of Concrete Research, Vol. 43, No. 157, Dec.
1991, pp. 219-231.
120. Dhir, R. K. and Byars, E. A., "Pulverised Fuel Ash Concrete: Permeation
Properties of Cover to Steel Reinforcement", Cement and Concrete Research,
Vol. 23, No.3, 1993, pp. 554-556.
240
References 2-11
122.Nixon, P. 1., Hardcastle, 1. and Ben-Bassat M., "The effect of differences in the
composition of Portland Cement on the Properties of Hardened Concrete",
Magazine of Concrete Research, Vol. 42, No. 151, Jun. 1990, pp. 59-66.
2·B